Interactive Brokers Group IBKR

Reference table

1. What the business is, and the two definitions the vendor gets wrong

Interactive Brokers Group, Inc., Greenwich, Connecticut; CIK 0001381197; NASDAQ-listed since 2007; chief executive Milan Galik; founder and chairman Thomas Peterffy; 3,182 employees. The 10-Q's self-description: "We are an automated global broker... routing orders and executing, clearing and settling trades in stocks, options, futures, forex, bonds, mutual funds, ETFs and precious metals on more than 170 electronic exchanges and market centers in 40 countries and 29 currencies... Our customer base is diverse... Currently, our customers reside in over 200 countries and territories." It also notes: "We also offer trading in prediction markets, which are event-based contracts traded on ForecastEx, a CFTC-registered exchange and clearinghouse we established."

Correction 1 — the revenue is a gross-up

From the 8-K of 2026-07-21, verbatim, in the company's own format (US$M):

The vendor's inc_q.revenue for the same quarter is $2,953M — because it reports gross interest income and pushes interest expense into totalOtherIncomeExpensesNet. For the full year 2025 the vendor reports $10,232M where the company's four quarters of total net revenues sum to $6,120M.

And the file contradicts itself, which is how the defect is proved. est.revenueAvg for the year ended 2025-12-31 is $6,121.7M — a 0.03% match to the company's net figure and 40% below the vendor's own income statement. earn_cal.revenueActual for the June 2026 quarter is $1,896,000,000 — the filed net figure, exactly.

> The vendor reports 2025 revenue of $10,232M; the company reports $6,120M of total net revenues; the vendor's own estimate block and earnings calendar both use the company's figure. We use the net figure throughout, and no revenue growth rate or price-to-sales ratio in this dive is computed on the gross-up.

This is the SCHW defect from the defect register, and it is the second confirmed instance.

Correction 2 — the share count and what the public actually owns

From the 10-Q filed 2026-05-07, verbatim:

> "Interactive Brokers Group, Inc. (the "Company" or "IBG, Inc.") is a holding company whose primary asset is its ownership of approximately 26.3% of the membership interests of IBG LLC. The remaining approximately 73.7% of IBG LLC membership interests are held by IBG Holdings LLC ("Holdings"), a holding company that is owned by our founder and Chairman, Mr. Thomas Peterffy and his affiliates, management and other employees of IBG LLC, and certain other members."

And the June quarter's income statement shows the consequence:

The vendor's marketCap of $153,155,514,803 divided by $88.63 implies 1,728.0 million shares — the fully-exchanged figure, close to but 1.9% above the 1,696.4 million the filing reports. Every per-share ratio in ratios_ttm and km_ttm uses roughly 446 to 450 million (verified: revenuePerShareTTM of $24.798 against trailing gross revenue of $11,107M gives 448.0M).

Why the price-to-earnings ratio survives and nothing else does. The share price is a Class A price and the earnings per share is computed on the Class A share of earnings, so $88.63 ÷ $2.51 = 35.3x is internally consistent and correct. But priceToSalesRatioTTM of 13.79x is the fully-exchanged market capitalisation of $153.2B divided by the GROSS-UP trailing revenue of $11,107M — wrong on both the numerator and the denominator. The internally consistent figure is fully-exchanged equity value of $150.3B (1,696.4M × $88.63) divided by trailing NET revenues of $6,778M = 22.2x.

2. The June quarter — and the operating leverage underneath it

Non-interest expenses, from the earnings release (US$M):

Three observations.

First, the cost base is almost fixed and almost trivially small. $440 million of quarterly expense supports $1,896 million of net revenue. Employee compensation of $182 million for 3,182 people is roughly $229,000 a head annualised — competitive pay for a technology firm — and it grew 12% against 28.1% revenue growth. Communications expense was $11 million, flat. This is why the pre-tax margin expands whenever revenue grows, and it is the entire operating-leverage case.

Second, the one line that grew faster than everything else is execution, clearing and distribution fees, up 22%, and the release explains why: "driven by a $19 million increase in regulatory fees, as the SEC Section 31 transaction fee rate increased on April 4, 2026; partially offset by greater capture of liquidity rebates from certain exchanges due to higher trading volumes." A regulatory pass-through, not a cost problem.

Third, and this is the only genuinely negative number in the release: customer bad debt expense rose from $1 million to $10 million year on year, and from $2 million to $11 million for the half. In absolute terms it is 0.5% of quarterly net revenues and irrelevant. In relative terms it is a tenfold increase and it is the first credit signal anywhere in this file, on a firm whose knowledge-base case rests partly on "fully-automated margin liquidation" letting it "underprice margin loans while carrying less real risk than peers." We flag it as a number to watch rather than as a problem.

The revenue mix and what drives it

The release attributes the commission growth to volume, not to price: "Commission revenue increased 30% to $673 million on higher customer trading volumes. Customer trading volume in options, stocks and futures increased 17%, 14% and 2%, respectively." Commissions grew 30% on 17% options-volume growth, so mix and product also contributed.

And it attributes net interest income to balances, not to rates: "Net interest income increased 23% to $1.06 billion primarily on higher average customer margin loans and customer credit balances." That is an important distinction for the rate-sensitivity question in Section 5, and it is the company's own framing, so it is half-weighted.

The multi-year record, on the company's own basis

(Net revenues derived as vendor revenue less interestExpense; the 2025 figure is confirmed against the sum of the four quarterly filed figures — $1,427M + $1,480M + $1,570M + $1,643M = $6,120M — and against est.revenueAvg of $6,121.7M.)

Net revenues have compounded at approximately 23.2% a year over five years and the pre-tax margin has risen from roughly 58% to 78%. That is the whole company in one table, and none of it is visible in the vendor's revenue series.

3. Balance sheet — a broker's, and no ratio in the file survives it

At 2025-12-31 (the most recent annual balance sheet in the payload; the June-quarter 10-Q had not been filed as of this dive, and the earnings release reports total equity of $22.3 billion at 2026-06-30):

Almost every ratio computed on this balance sheet is meaningless and we reject them explicitly rather than quietly:

What can be said about the balance sheet is that it is over-capitalised by design. The June earnings release states total equity of $22.3 billion, and our knowledge base supplies the regulatory-capital framing at conviction 76: "Fortress balance sheet with ~$18B excess capital (~95% of total) at end-2025 allows serene growth independent of market cycles." That is a knowledge-base claim rather than a filing figure and it is labelled as such, but it is directionally consistent with a broker carrying $19 million of debt against $22.3 billion of equity.

debtToAssetsRatioTTM of 18.4% and interestCoverageRatioTTM of 2.17x are computed on the gross interest expense of a brokerage funding book and describe nothing about solvency. Interactive Brokers has $19 million of borrowings.

4. Cash flow and capital returns

Free cash flow is not a meaningful concept for a broker either, and the vendor's figures show why: freeCashFlowYieldTTM reads 10.9% and priceToOperatingCashFlowRatioTTM reads 2.36x, because operating cash flow for a broker includes movements in customer balances and segregated funds. operatingCashFlowSalesRatioTTM of 151.6% — cash flow greater than revenue — is the tell. We do not use any cash-flow multiple in this dive.

Capital returns are modest and that is itself part of the case. dividendPerShareTTM is $0.3275, a 0.37% yield; profile.lastDiv is null. Interactive Brokers has paid roughly $0.08 a quarter since the four-for-one split of June 2025 (the 10-Q: "Each holder of record of common stock as of the close of market on June 16, 2025, received three additional shares"), and dividendPayoutRatioTTM of 12.7% on trailing EPS of $2.51 is consistent with that. stockBasedCompensationToRevenueTTM of 1.08% is among the lowest in this batch.

Our knowledge base names the capital return as an unrealised catalyst. we_study_billionaires (2024-01-11, conviction 55): "Catalysts: dividend increase or big buyback on lazy balance sheet, or 79-year-old founder (owns 68%) eventually selling the company." Two and a half years later the dividend is still 0.37% and there is no buyback in the file. We record the claim as unrealised rather than wrong, and note that the ownership figure in it (68%) is close to but below the 73.7% the 10-Q reports for IBG Holdings LLC as a whole.

5. Valuation — priced in or room?

At $88.63, Class A shares 445,616,326, fully-exchanged units 1,696,353,742 (fully-exchanged equity value $150.35B):

Estimate coverage is the thinnest in this batch — FIVE analysts on 2026 and 2027 revenue, SIX and SEVEN on EPS, FOUR on 2028, and TWO on 2029. The 2029 row is excluded from every conclusion and the 2028 row is used only as a cross-check. On a $150 billion company, five-analyst revenue coverage is itself a finding, and it is consistent with the 19-analyst rating distribution.

est.ebitdaAvg and est.ebitAvg trip the published ebitAvg > ebitdaAvg signature in two years — NOT USED. For 2023, ebitdaAvg of $722.2M against ebitAvg of $1,761.3M; for 2022, $656.6M against $1,601.2M. EBIT above EBITDA is arithmetically impossible. From 2024 onward the ratios normalise (EBITDA at roughly 80.9% of net revenues in every forward year — itself an effectively fixed ratio), but the rows are rejected in full and all forward valuation runs on epsAvg and revenueAvg.

Peer context. The vendor peer set is a mixed financials basket — Bank of America ($446.4B), UBS ($175.2B), BBVA ($155.4B), Chubb ($134.4B), ICICI Bank ($108.5B), KKR ($97.2B), CME ($94.4B), Robinhood ($84.1B), Intercontinental Exchange ($83.8B), Brookfield ($99.7B). Robinhood is the only true retail-brokerage comparable and CME and ICE are the exchange comparables. No peer multiples are supplied in the file, so no peer-multiple comparison is drawn. One knowledge-base claim bears on the comparison directly: compound_and_friends (2026-06-23, conviction 62) — "Prefer the brokers over the exchanges right now."

5a. What today's price assumes (the inversion)

At $88.63 — 32.5x the 2026 consensus and 27.8x 2027 — the price embeds the following falsifiable claims, each with a number and a date:

• Diluted EPS of $2.724 in 2026 and $3.186 in 2027. (Consensus-derived; 6 and 7 analysts.) The first half delivered $1.29 ($0.60 + $0.69), so the second half needs $1.43. The nearest test is 2026-10-20 against $0.702, which is barely above the $0.69 just delivered — consensus is asking for almost no sequential progress on a business that grew net revenues 28% year on year. That is a low bar and it is worth naming as one.
• Net revenues of $7,491.9M in 2026 and $8,677.9M in 2027 — +22.4% and +15.8%. (Consensus-derived; 5 analysts each.) The deceleration to 15.8% is the assumption that carries the most weight in the 2027 number and it sits against a quarter that just delivered 28.1%.
• The pre-tax margin holds at or above 77%. (Filing-confirmed; our assumption on the hold.) It went 75% to 77% in a year and has risen from roughly 58% in 2020. With $440 million of quarterly cost against $1,896 million of revenue there is very little left to squeeze, so further margin expansion has to come from revenue growth — which means the margin assumption and the revenue assumption are the same assumption.
• Net interest income, at 55.7% of net revenues, does not fall. (Filing-confirmed share; our assumption on the direction.) This is the most fragile assumption in the price. The company attributes the growth to "higher average customer margin loans and customer credit balances" rather than to rates, and our knowledge base argues at conviction 78 that the market overstates the rate sensitivity — "only ~10% NIM decline per 100bps cut." But 55.7% of revenue is 55.7% of revenue, and neither the filings nor the payload contains a rate-sensitivity disclosure we can test it against.
• The market keeps paying 28-33x forward earnings. (Our number.) At 20x the 2027 consensus the stock is $64 — above the 52-week low; at 36x it is $115. Interactive Brokers traded on a "reasonable ~15 PE" as recently as the January 2024 knowledge-base claim in our own lane, so the multiple has more than doubled and the re-rating is a large part of the twelve-month return.

5b. The return bridge (why the multiple moves)

Expected return over the next twelve months decomposes as: diluted EPS growth (+17.0%, from 2026E $2.724 to 2027E $3.186) + multiple drift (COMPRESSION, from 32.5x on the current forward year to roughly 27.2x on the then-forward year, −16%) + dividend yield (+0.37%) ≈ +11% to +13%.

Our base assumes compression and the reason is the estimate path rather than a de-rating call. Today the market pays 32.5x on the current year and 27.8x on the next. Our $99 is 31.1x the 2027 consensus and 27.2x the 2028 consensus — so in twelve months, when 2028 is the forward year, the base implies 27.2x against today's 27.8x on the equivalent horizon. That is close to flat and it is a deliberate choice: we assume the multiple broadly holds and the return comes from earnings.

The sensitivity is the real story. If the multiple HELD at 32.5x on the 2028 consensus the price would be $118 (+33%); at 24x, $87 — today's price; at 20x, $73 (−18%). Almost the entire dispersion is a multiple that has roughly doubled in two and a half years, on a name with 19 analysts covering it.

The bull case at $127 requires the multiple to stay near where it is (40.2x the 2027 consensus, 34.9x the 2028) and the growth not to decelerate. The bear case at $64 is 20.1x the 2027 consensus — the multiple the company carried before the re-rating — and it sits 6% above the 52-week low.

5c. Variant perception (where we differ, what would surprise)

• We differ from every screen on what this company's revenue is and what it is worth per unit of it. The published revenue of $10,232M is a gross-up; the company's figure is $6,120M. The published priceToSalesRatioTTM of 13.79x divides a fully-exchanged market capitalisation by that gross-up; the internally consistent figure is 22.2x on net revenues, which is a materially more demanding number and it cuts against the position. We report it for that reason.
• We differ on the share-count arithmetic, and here the correction is neutral rather than favourable. The market capitalisation implies 1,728.0 million shares against a filed fully-exchanged 1,696.4 million and a traded Class A 445.6 million. The price-to-earnings ratio is unaffected because price and EPS are both struck on the Class A slice — but every reader should know that owning IBKR is owning 26.3% of IBG LLC alongside a founder-controlled vehicle that holds 73.7% and all of the Class B voting stock.
• We differ from the knowledge base on how settled the rate question is, and this is our substantive edge on the name. The lane's most specific claim (business_breakdowns, conviction 78) is that "the market misunderstands rate sensitivity: paying base-rate-minus-50bps means only ~10% NIM decline per 100bps cut, dwarfed by 30-35% account growth." We cannot verify it: neither the 10-Q nor the earnings release in this archive contains a rate-sensitivity disclosure, and the vendor payload has none. It is a plausible, specific, unverified claim about 55.7% of the revenue, and we treat it as a hypothesis to be tested rather than as a fact underwriting the multiple. Watchable event: net interest income in the first quarter in which policy rates fall.
• Positive surprise that would force a re-rate: third-quarter net revenues materially above $1,932.9M with the pre-tax margin at or above 77% — which, given the quarter just delivered $1,896M and consensus asks for +1.9% sequentially, is a live possibility rather than a stretch. A second, more consequential one: a material dividend increase or a buyback, which the 2024-01-11 knowledge-base claim named as a catalyst and which has still not happened on a balance sheet carrying roughly $22.3 billion of equity.
• Negative surprise that would break the thesis: net interest income falling year on year; customer bad debt continuing to compound from the $11 million first-half figure; commission growth decelerating below 15%; the pre-tax margin falling below 74%; or any change in the Class B control arrangement or in Thomas Peterffy's position that public shareholders do not control and cannot anticipate.

Synthos fair values

All three anchors are multiples of the 2027 consensus diluted EPS distribution (mean $3.186, low $3.005, high $3.296, 7 analysts). The distribution is unusually tight — the high is only 9.7% above the low — which reflects thin coverage rather than genuine agreement, and we say so.

• Bear ~$64 — 20.1x the 2027 consensus mean, cross-checked at 21.3x the consensus LOW of $3.005; roughly 6% above the 52-week low of $60.16, and close to the multiple the company carried before its re-rating. The scenario: policy rates fall and the 55.7% of net revenues that is net interest income proves more sensitive than the knowledge base claims; account and volume growth decelerate; customer bad debt keeps compounding; and a 28x multiple becomes a 20x multiple. −27.8%.
• Base ~$99 — 31.1x the 2027 consensus MEAN of $3.186, equivalently 27.2x the 2028 consensus mean of $3.639 and 36.3x the 2026 consensus. Cross-check: fully-exchanged equity value of $167.9B on 2027 consensus net revenues of $8.678B is 19.3x sales. Sensitivity, stated openly: 28x the 2027 consensus gives $89 — essentially spot — and 34x gives $108, essentially the street. The answer lives in a 28-34x band. The scenario: net revenue growth moderates from 28% toward the high teens, the pre-tax margin holds near 77%, account growth continues, and the multiple is broadly unchanged. +11.7%, plus a negligible dividend.
• Bull ~$127 — 39.9x the 2027 consensus mean, cross-checked at 34.9x the 2028 mean and 38.5x the consensus high. The scenario: net revenue growth stays near 25%, ForecastEx and crypto convert from products into revenue lines, the balance-sheet catalyst arrives as a material dividend increase or buyback, and the market pays a growth multiple for a business with a 77% pre-tax margin. +43.3%.

Base is 11.7% above spot; asymmetry approximately 1.56:1 to the upside (27.8% down, 43.3% up). That ratio, plus an oversold technical, a fifth consecutive beat, a street whose lowest target is 10.6% above spot, and recurring open-market director buying, is enough for Tactical. It is not enough for Core with 55.7% of revenue in an untested rate exposure and a public float owning 26.3% of the economics.

6. Knowledge base — sixteen claims, none bearish, and nine from one recording

Raw hits: 16 (entity matches). Discarded as off-topic: 1. Quarantined: 1. Used: 14. Channels: 7.

This is the richest and the most one-sided lane in this batch — sixteen claims and not one bearish — which is precisely why it needs the hardest scrutiny.

The concentration test, and it is severe

business_breakdowns supplies 9 of 16 raw hits — 56.3% — and ALL NINE carry the same date, 2025-05-17. Nine claims distilled from one episode of one show, fifteen months before this dive.

Stripping business_breakdowns leaves 7 claims. One (natalie_brunell, 2025-09-02, on the Genius Act and stablecoins) does not mention Interactive Brokers in its thesis and is discarded as off-topic. One is quarantined. That leaves FIVE usable claims from FIVE independent channels — and all five are bullish.

> Result: the lane SURVIVES its concentration in direction while losing most of its substance. Unlike Deere, where stripping the dominant channel collapsed the lane, here the remaining five claims all point the same way. But every specific, mechanical, structural argument about why Interactive Brokers is a good business comes from a single recording, and that is a limitation we cannot resolve from inside this file.

The single-episode lane — quoted, because the arguments are good

> 2025-05-17 · bullish · conviction 88 · business_breakdowns · no named speaker

> "Lowest-cost, most-automated broker with highest quality; flywheel compounds account growth without marketing, driving durable share gains."

> 2025-05-17 · bullish · conviction 85

> "Account growth, not trades or revenue-per-account, is the true value driver; per-account economics are stable so growth compounds intrinsic value."

> 2025-05-17 · bullish · conviction 82

> "Founder-driven risk conservatism plus fully-automated margin liquidation lets them underprice margin loans while carrying less real risk than peers."

> 2025-05-17 · bullish · conviction 80

> "Direct market access instead of payment-for-order-flow gives best execution and a structural cost advantage competitors can't replicate."

> 2025-05-17 · bullish · conviction 80

> "Growth doubling via B2B channels—RIAs, hedge funds (now 5th-largest prime broker), and introducing brokers like HSBC renting the tech stack."

> 2025-05-17 · bullish · conviction 78 — the most useful and most testable claim in the lane

> "Market misunderstands rate sensitivity: paying base-rate-minus-50bps means only ~10% NIM decline per 100bps cut, dwarfed by 30-35% account growth."

> 2025-05-17 · bullish · conviction 76

> "Fortress balance sheet with ~$18B excess capital (~95% of total) at end-2025 allows serene growth independent of market cycles."

> 2025-05-17 · bullish · conviction 75

> "Counter-positioning makes M&A uneconomic (acquired revenue collapses at their spreads), so growth stays organic and competitors fold rather than compete."

> 2025-05-17 · bullish · conviction 65

> "Long runway: 3.5M accounts vs Schwab/Fidelity ~25M; targets of 20M then 80M as active traders graduate from Robinhood/eToro."

What Interactive Brokers' own disclosure corroborates and what it does not. The cost claim is corroborated from a different direction: 3,182 employees, $440 million of quarterly non-interest expense against $1,896 million of net revenues, a 77% pre-tax margin, and $11 million of quarterly communications expense. The balance-sheet claim is directionally consistent with $22.3 billion of total equity against $19 million of borrowings. The rate-sensitivity claim, the account-growth targets and the business-to-business figures are NOT corroborated by anything in the archive or the payload, and are recorded as unverified.

One further finding from the same channel, retrieved by a case-sensitive sweep rather than the entity search: business_breakdowns (2025-05-17, bearish on Charles Schwab, conviction 55) — "Schwab's non-payment on customer cash caused a funding run when rates rose, contrasting IBKR's honest transparent deposit arrangement." A competitor-negative claim that is Interactive-Brokers-positive by construction, and it is the mechanism behind the "paying base-rate-minus-50bps" argument above.

The five claims that survive the concentration test

> 2024-01-11 · bullish · conviction 80 · we_study_billionaires · no named speaker

> "20%+ account growth for years; lowest-cost global platform, high switching costs, pristine balance sheet, reasonable ~15 PE; huge international runway where zero-commission is illegal."

The business argument stands and the valuation premise has expired. The claim's explicit anchor is a "reasonable ~15 PE"; the stock now trades at 35.3x trailing and 27.8x the 2027 consensus. We record that plainly: the thesis worked, and the reason it worked was partly a doubling of the multiple that is not available a second time. The international point — that zero-commission models are illegal in many non-US markets — is the runway argument and remains untested here.

> 2024-01-11 · bullish · conviction 55 · we_study_billionaires

> "Catalysts: dividend increase or big buyback on lazy balance sheet, or 79-year-old founder (owns 68%) eventually selling the company."

Two and a half years on, neither catalyst has occurred: the dividend yield is 0.37% and there is no buyback in the file. The ownership figure of 68% is close to the 10-Q's 73.7% for IBG Holdings LLC in aggregate.

> 2024-10-04 · bullish · conviction 65 · empire · no named speaker

> "Prediction markets are gaining legitimacy — Kalshi's legal win plus Interactive Brokers launching presidential betting signal mainstream adoption."

Corroborated by the 10-Q, which discloses ForecastEx as "a CFTC-registered exchange and clearinghouse we established."

> 2026-04-01 · bullish · conviction 50 · anthony_pompliano_show · SPEAKER: Matthew Siegel · speaker_role: independent

> "Owns Interactive Brokers — you can now send crypto to your IBKR account (a first for them), and their 12-bip trading fees are attractive."

A named guest speaker with a disclosed position. Per the house rule on this channel, the host's own voice would be low-weighted; Siegel is a guest and the distiller records his speaker_role as "independent", so he is treated as first-class. Conviction 50 is modest and the claim is a product observation rather than a thesis.

> 2026-06-23 · bullish · conviction 62 · compound_and_friends · no named speaker · TEST: "IBKR breaks $100"

> "Prefer the brokers over the exchanges right now; Interactive Brokers made a record high today and is going to break 100."

> The test attached to this claim has NOT been met. The 52-week high in the technical block is $97.41 and the stock closed $88.63 on 2026-08-04, six weeks later. We report the failure rather than counting the claim as support. The relative-preference half — brokers over exchanges — is not testable from this file.

The management claim — excluded from the pool, quoted at half weight

> 2026-04-09 · bullish · conviction 80 · odd_lots · SPEAKER: THOMAS PETERFFY · speaker_role: MANAGEMENT

> "Interactive Brokers' Forecast Trader is currently a loss-leader, but Peterffy is very sure it will eventually become very profitable, backed by IBKR's serious institutional client base."

This is the founder and chairman of the company speaking about his own company's product, correctly tagged by the distiller as management voice. Per the standing house rule it is EXCLUDED from the conviction pool and quoted here separately at half weight. Note what it actually says, which is more useful than a promotional line would be: Forecast Trader is a LOSS-LEADER TODAY. That is a management admission against interest, and it is the best available answer to whether prediction markets are currently a revenue line. They are not.

Excluded

One claim (2023-07-18, conviction 78, skill 2.0) carries the quarantine_misattributed thinker slug with an underlying doc_id pointing at a jordi_visser recording, and asserts that Coinbase's monthly active users exceed those of Interactive Brokers, Schwab and Fidelity. Excluded outright under the attribution policy, and it is in any case a claim about Coinbase. A second claim (bankless, 2026-03-16, conviction 65) names ForecastEx as an entity within an argument about Robinhood's prediction-market strategy; it is a competitor claim rather than a view on Interactive Brokers and is not counted.

Conclusion. The Synthos knowledge base is unanimously positive on Interactive Brokers and the depth of that positivity rests on one recording. The claims that survive independent sourcing are five, all bullish, mostly low-to-medium conviction, one from management and half-weighted, and one carrying a test that has failed. Conviction rating: Medium. Breadth: 7 nominal, 5 effective. Net conviction: positive, concentrated, with one failed test. A modest positive weight is applied to the verdict tier and no premium is applied to the fair value.

7. Data integrity — what we rejected from the vendor file and why

Eleven findings. IBKR is the most structurally misdescribed file in this batch: the revenue definition, the share count and every balance-sheet ratio are all wrong for the same underlying reason — the vendor is applying an industrial template to a broker held through a partnership.

1. revenue is a GROSS-UP the company never publishes — REJECTED and replaced. inc_a.revenue for 2025 is $10,232M, being total revenues including gross interest income of roughly $7.3 billion, with interest expense of $4,219M pushed into totalOtherIncomeExpensesNet. The company's own top line is "Total net revenues" of $6,120M (the four filed quarterly figures: $1,427M + $1,480M + $1,570M + $1,643M). The file contradicts itself: est.revenueAvg for 2025 is $6,121.7M — a 0.03% match to the NET figure — and earn_cal.revenueActual for the June 2026 quarter is $1,896,000,000, the filed net figure exactly. No revenue growth rate or sales multiple in this dive is computed on the gross-up. This is the second confirmed instance of the SCHW defect class.

2. marketCap implies 1,728.0M shares against a filed Class A count of 445,616,326 — REJECTED for every purpose except the P/E. The 10-Q states IBG, Inc. owns approximately 26.3% of IBG LLC through 445,616,326 membership interests, with IBG Holdings LLC holding 1,250,737,416, total 1,696,353,742. Every per-share ratio in the payload uses 446-450M (revenuePerShareTTM of $24.798 against $11,107M of trailing gross revenue implies 448.0M). The price-to-earnings ratio of 35.3x survives because price and EPS are both Class A figures; priceToSalesRatioTTM of 13.79x does not, because it divides the fully-exchanged market capitalisation by the gross-up revenue — wrong in the numerator and the denominator simultaneously. The internally consistent figure is 22.2x net revenues on a $150.35B fully-exchanged equity value.

3. Every liquidity, leverage and turnover ratio is a broker-balance-sheet artefact — REJECTED as lenses. currentRatioTTM and quickRatioTTM 17,791.6x; cashRatioTTM 4,946.5x; operatingCashFlowRatioTTM 1,295.2x; daysOfSalesOutstandingTTM 4,077 days; financialLeverageRatioTTM 41.9x; debtToEquityRatioTTM 7.69x against $19M of actual borrowings; returnOnAssetsTTM 0.46% on $203 billion of mostly customer assets. None describes anything about this company.

4. enterpriseValueTTM, evToSalesTTM and evToEBITDATTM are not meaningful for a broker — REJECTED. $134.3B, 12.09x and 13.65x are computed by netting customer-related cash against a fully-exchanged market capitalisation. Enterprise value is not a construct that applies to a firm whose balance sheet is its customers' money.

5. est.ebitdaAvg and est.ebitAvg trip the published ebitAvg > ebitdaAvg signature — NOT USED. For 2023, ebitdaAvg of $722.2M against ebitAvg of $1,761.3M; for 2022, $656.6M against $1,601.2M. EBIT above EBITDA is impossible. From 2024 onward the EBITDA margin settles at an effectively constant 80.9% of revenueAvg in every forward year, which is the fixed-ratio signature. All forward valuation runs on epsAvg and revenueAvg.

6. epsdiluted is NULL on every row and eps is the Class A basic figure. The vendor's 2025 eps of 2.23 is on the Class A slice; the earnings release reports diluted EPS of $0.69 for the June quarter, both reported and adjusted — Interactive Brokers is one of very few companies in this batch where reported and adjusted EPS are identical, which is worth recording. All EPS figures here are filing-sourced.

7. seg_prod and seg_geo are BOTH EMPTY. Interactive Brokers reports revenue in four lines (commissions, other fees and services, other income, net interest income) and none is in the payload. All revenue disaggregation comes from the 8-K of 2026-07-21.

8. effectiveTaxRateTTM of 8.59% is correct and highly misleading. IBG, Inc. pays corporate tax only on its 26.3% share of IBG LLC's income; the remaining 73.7% passes through to the members of IBG Holdings LLC and is taxed at their level. The June quarter shows $118M of tax on $1,456M of consolidated pre-tax income — an 8.1% consolidated rate — with $1,026M of the after-tax result going to noncontrolling interests. taxBurdenTTM of 0.210 is not a tax rate at all; it is net income attributable divided by consolidated pre-tax income, which is the ownership ratio wearing a tax label.

9. earn_cal.revenueActual is on the correct NET basis but does not always match the filing exactly. For the June 2026 quarter it reports $1,896,000,000 against the filed $1,896M — exact. For the March 2026 quarter it reports $1,680,000,000 against the filed $1,669M — a $11M (0.7%) gap. For the June 2025 quarter it reports $1,438,000,000 against the $1,480M the Q2 2026 release gives as the prior-year comparative — a $42M (2.8%) gap. We use the filings for every period.

10. All cash-flow multiples are meaningless and are not used. freeCashFlowYieldTTM of 10.9%, priceToOperatingCashFlowRatioTTM of 2.36x and operatingCashFlowSalesRatioTTM of 151.6% — operating cash flow greater than revenue — all reflect movements in customer balances and segregated funds running through operating activities.

11. lastDividend is absent and dividendPerShareTTM of $0.3275 is the trailing total. profile.lastDiv is null. The quarterly rate has been approximately $0.08 since the four-for-one forward split effective for holders of record 2025-06-16, which the 10-Q documents in full and which the vendor has handled correctly — all prior-period per-share figures in the payload are split-adjusted, verified against the 10-Q's statement that "All prior period shares, per share amounts and stock incentive awards presented herein have been retroactively adjusted to reflect the stock split." No split-adjustment defect is present, which after the KLAC precedent is worth stating.

Not defects, correctly reported: priceToEarningsRatioTTM of 35.03x is internally consistent (Class A price ÷ Class A EPS) and is the one headline multiple in this file that survives; netIncomePerShareTTM of $2.514 reconciles to the four filed quarterly EPS figures ($0.57 + $0.65 + $0.60 + $0.69 = $2.51); dividendPayoutRatioTTM of 12.7% is arithmetically correct; stockBasedCompensationToRevenueTTM of 1.08% is accurate and genuinely low; and the split adjustment is clean throughout.

Non-equity tripwire — checked and passed, with a control note. IBKR is Class A common stock, NASDAQ-listed since 2007-05-04, beta 1.329, a variable dividend, a 52-week range of $60.16 to $97.41 (61.9%), volume of 4.30 million shares (roughly $381M of turnover). The control structure is disclosed rather than a defect: Holdings owns all of IBG, Inc.'s Class B common stock, which the 10-Q states "has voting rights in proportion to its ownership interests in IBG LLC" — so IBG Holdings LLC controls approximately 73.7% of the votes and public Class A holders can never outvote it.

8. Technicals

• Price $88.63. −9.0% from the 52-week high of $97.41; +47.3% above the 52-week low of $60.16. Position within the annual range: 76th percentile.
• The moving averages straddle: −2.0% below the 50-day at $90.46, +16.3% above the 200-day at $76.20, with the 50-day well above the 200-day. A strong uptrend in a short-term pullback — the cleanest "buy the dip" configuration in this batch.
• RSI 34.2 — the ONLY oversold reading among these twelve names, and the second-lowest after nothing else comes close. MACD −0.75 — negative, consistent with a pullback rather than a breakdown.
• Maximum drawdown from peak over the trailing year: −9.0%, i.e. this stock has not had a correction worth the name.
• Relative performance is the strongest in this batch on every horizon: 3-month +8.5% vs SPY +7.6% and QQQ +7.7%; 6-month +17.7% vs SPY +11.1% and QQQ +15.7%; 12-month +39.3% vs SPY +24.3% and QQQ +30.8% — a 15-point outperformance against the index and 8.5 points against the Nasdaq.
• Sentiment: 10 buy, 8 hold, 1 sell across NINETEEN analysts — the thinnest coverage in this batch on a $150 billion company. Consensus $107.75 (+21.6%), median $109.50 (+23.5%), high $114 (+28.6%), low $98 — 10.6% ABOVE spot. The entire target range is above the price and spans only 16%, which on 19 analysts we read as limited independent work rather than as strong agreement.

Today's move

IBKR closed 2026-08-04 at $88.63, up $0.88 or 1.00% from $87.75. It opened at $88.75, traded $86.92 to $89.48, and closed mid-range on 4,296,783 shares — roughly $381 million of turnover. An unremarkable session. The one piece of company-related news dated 2026-08-04 is in the insider file: a Form 4 filed that day recording a director's open-market purchase on 2026-08-03.

9. Insiders — a director buying every month, and a tax-withholding event

Eight transactions:

The reading, in two parts.

The F-InKind block is not selling. All five entries are dated 2026-05-08 at the same price of $84.42 and are shares withheld by the issuer to satisfy tax on vesting awards. The employee never receives them and never sells them. Four officers and one accounting officer, one date, one price, is the signature of a scheduled annual vest.

The purchases are small, genuine and systematic. Director Lori A. Conkling bought 25 shares in the open market on 2026-06-01 at $86.10, 25 on 2026-07-01 at $88.03, and 25 on 2026-08-03 at $87.03 — the last filed on 2026-08-04, the date of this dive. Seventy-five shares across three months is approximately $6,500 of stock, which is economically trivial and has the regular monthly cadence of an automatic purchase plan rather than a discretionary conviction trade. We record it as a genuine P-Purchase pattern and assign it minimal weight.

What the file does contain, and it is the more meaningful figure, is the size of the executive holdings. Chief executive Milan Galik holds 3,215,389 shares — $285.0 million at the closing price. Chief financial officer Paul Brody holds 2,856,806 — $253.2 million. Those are enormous unhedged personal stakes in a company whose founder controls 73.7% of the economics, and their persistence is a mild positive. There is no open-market sale anywhere in this file.

10. Verdict, kill-criteria and flip conditions

Buy — Tactical.

What is genuinely exceptional, and all of it is filing-verified: total net revenues of $1,896 million, up 28.1%, with commissions +30%, net interest income +23%, other fees +40% and other income +88% simultaneously; diluted EPS of $0.69, up 35.3%; a pre-tax margin of 77%, up from 75%, on total non-interest expenses of $440 million and 3,182 employees managing $203 billion of assets; total non-interest expenses growing 17.0% against 28.1% revenue growth; $22.3 billion of equity against $19 million of borrowings; a fifth consecutive earnings beat; RSI at 34.2 with the price 16.3% above a rising 200-day moving average; and a knowledge-base lane of sixteen claims with none bearish.

What keeps this Tactical rather than Core: 55.7% of net revenues is net interest income whose rate sensitivity our own research calls misunderstood but which no source in the contract lets us verify; 27.8x the 2027 consensus after a 39.3% twelve-month advance, against a knowledge-base claim from January 2024 whose explicit premise was a "reasonable ~15 PE"; public shareholders own 26.3% of IBG LLC and received $312 million of the $1,338 million the group earned last quarter, with IBG Holdings LLC controlling the votes; customer bad debt expense up tenfold year on year, small but new; nineteen analysts covering a $150 billion company and five analysts on the revenue estimate; and a knowledge-base lane in which nine of sixteen claims come from a single recording and one carries a test that has failed.

The distinction that matters. The quality here is not in dispute and the data describing it is almost entirely wrong. A screen reading this file sees $10.2 billion of revenue (70% too high), 1.73 billion shares (against 446 million traded), a 13.79x price-to-sales ratio (wrong twice) and a 17,791x current ratio. What the filings actually show is one of the most operationally efficient businesses in public markets, growing net revenues at 28% with a 77% pre-tax margin, at 27.8x forward earnings, in a technical pullback. That is a position — sized for the fact that more than half the revenue depends on an interest-rate exposure nobody in this file has quantified, and that we would be buying a quarter of a company whose founder owns the rest.

Pre-registered KILL criteria — what would take this to Watch or Avoid:

• Net interest income falling year on year. It is 55.7% of net revenues and the single largest dependency in the thesis.
• Pre-tax margin below 74%. It is 77% and rising, and there is very little cost left to cut if revenue stalls.
• Commission growth below 15%, from 30%, which would say the volume flywheel has slowed.
• Customer bad debt above roughly $25 million in a quarter, from $10 million — the credit signal compounding rather than normalising.
• Any change in the Class B control arrangement or in IBG Holdings LLC's position that public shareholders learn about after the fact.
• Consensus 2027 net revenues revised below $8.0 billion, which would confirm the deceleration to 15.8% is real and getting worse.

Pre-registered FLIP conditions — what would take this to Buy — Core:

• A disclosed rate-sensitivity analysis, or a quarter in which policy rates fall and net interest income holds. This single item is worth a full tier, because it is the only assumption in the price we cannot currently test and it governs 55.7% of the revenue.
• A material dividend increase or an inaugural buyback — named as a catalyst in our knowledge base in January 2024 and still unrealised on $22.3 billion of equity with a 0.37% yield.
• ForecastEx or crypto appearing as a disclosed revenue line. The founder's own description today is "a loss-leader", half-weighted; a revenue line would convert an option into an asset.
• A price below approximately $78 (roughly 24.5x the 2027 consensus, near the 200-day moving average) with growth intact, which would put the base case at +27%.
• Independent corroboration of the structural case. Nine of sixteen knowledge-base claims come from one recording on one date; a second, unrelated source making the same account-growth and cost-advantage arguments would materially raise the conviction rating.

Where IBKR fits in the Synthos Framework Portfolio. The financial-infrastructure sleeve, at 0% today with a 2-3% target initiated here. On batch overlap: IBKR, BLK and BX are all financials in this batch but they are three different exposures — BlackRock is a fee on assets, Blackstone is a fee on private-market performance, and Interactive Brokers is a fee on transactions plus a spread on customer balances. Interactive Brokers is the only one of the three whose revenue rises when markets are volatile rather than when they are high, which makes it a genuine diversifier against the other two rather than a third helping of the same trade. Logged as a tracked Synthos call (Buy — Tactical) as of 2026-08-04 at $88.63.

Single biggest risk: net interest income is 55.7% of net revenues and nobody in this file has quantified its rate sensitivity. The company attributes the 23% growth to "higher average customer margin loans and customer credit balances" — volume, not rate — and our knowledge base argues at conviction 78 that the market overstates the exposure because Interactive Brokers pays customers base rate minus 50 basis points, implying only about a 10% net-interest-margin decline per 100 basis points of cuts. That claim is specific, plausible, unverified, fifteen months old, and comes from the same single recording that supplies nine of our sixteen claims. At 27.8x the 2027 consensus, with the pre-tax margin at 77% and almost no cost buffer, a rate-driven decline in the majority revenue line would go straight to earnings, and the multiple has no cushion in it.

Provenance & disclosures

• Traceability: 16 tagged knowledge-base claims name Interactive Brokers across 51,928 distilled claims — the largest lane in this batch, and ZERO of them are bearish, which is why it receives the hardest scrutiny here. 1 discarded as off-topic (natalie_brunell, 2025-09-02, on the Genius Act and stablecoins, which does not mention Interactive Brokers in its thesis); 1 quarantined (quarantine_misattributed, 2023-07-18, conviction 78, skill 2.0, underlying doc_id a jordi_visser recording, and in any case a claim about Coinbase); 14 used across 7 channels. Concentration sensitivity test run and reported: business_breakdowns supplies 9 of 16 raw hits (56.3%) and ALL NINE carry the same date, 2025-05-17 — nine claims from a single episode of one show. Stripping that channel leaves FIVE usable claims from FIVE independent channels, and all five are bullish, so the lane SURVIVES its concentration in direction while losing essentially all of its structural detail. The five are: we_study_billionaires ×2 (2024-01-11, conviction 80 and 55, no named speaker); empire (2024-10-04, conviction 65, no named speaker); anthony_pompliano_show with NAMED GUEST SPEAKER Matthew Siegel (2026-04-01, conviction 50, speaker_role: independent, position disclosed) — the house rule low-weighting that channel applies to the host, not to guests; and compound_and_friends (2026-06-23, conviction 62). TWO MANDATORY ADJUSTMENTS ARE APPLIED AND STATED. (i) odd_lots (2026-04-09, conviction 80) carries speaker: Thomas Peterffy and speaker_role: MANAGEMENT — the founder and chairman speaking about his own company — and is EXCLUDED from the conviction pool and quoted separately at half weight per house rule; note that its content is an admission against interest, that Forecast Trader "is currently a loss-leader." (ii) The compound_and_friends claim carries an explicit stored test — "IBKR breaks $100" — which has NOT been met: the 52-week high is $97.41 and the stock closed $88.63 six weeks later. The failure is reported rather than the claim being quietly counted. One further honesty point recorded in Section 6: the oldest surviving bullish claim rests explicitly on a "reasonable ~15 PE" and the stock now trades at 35.3x trailing, so its valuation premise has expired even though its business argument has not. All quotes are verbatim from the stored claim text.
• Data as-of: fundamentals — second-quarter 2026 income statement as at 2026-06-30, from the earnings release furnished as Exhibit 99.1 to the 8-K filed 2026-07-21 (22,625 characters, present in full with the complete statement of comprehensive income); balance sheet as at 2026-03-31 from the 10-Q filed 2026-05-07 — the second-quarter 10-Q had NOT been filed as of this dive, so no balance-sheet figure here is dated later than 2026-03-31 except the total-equity figure of $22.3 billion quoted directly from the earnings release; annual figures from the 10-K filed 2026-02-27 · estimates 2026-08-04 · prices 2026-08-04, quote timestamp 1785873601 = 2026-08-04T20:00:01Z ($88.63, +1.00%; 50-DMA $90.46; 200-DMA $76.20; RSI 34.2; MACD −0.75) · knowledge-base claims 2026-08-04. Fiscal year = calendar year. All figures come from the Synthos vendor data file for IBKR or from the SEC filings in the IBKR archive; no figure comes from memory, recall or external retrieval.
• Filing archive contents: 10-K filed 2026-02-27 (fiscal 2025, 462,032 characters, financial-statement tables preserved); 10-Q filed 2025-11-05 (September 2025 quarter); 10-Q filed 2026-05-07 (March 2026 quarter, 234,284 characters — the source for the ownership disclosure, the segment description and the balance sheet); 8-K filed 2026-07-21 carrying Exhibit 99.1, the full second-quarter 2026 earnings release with the complete income statement, financial highlights and pre-tax margin disclosure — the primary source for this dive; two 8-Ks filed 2026-07-31 (3,482 and 3,393 characters, cover documents only, contents not extracted). Note that the archive contains duplicate copies of several documents under both accession-suffixed and unsuffixed filenames; they are byte-identical and only one of each was read.
• Where the filings contradicted or corrected the vendor (detailed in Section 7): revenue identified as a GROSS-UP the company never publishes — $10,232M for 2025 against total net revenues of $6,120M, a 67% overstatement, with the vendor's own est.revenueAvg of $6,121.7M and earn_cal.revenueActual of $1,896M both matching the company's NET basis, so the file disagrees with itself by 40% on a definition it never labels; marketCap shown to imply 1,728.0M shares against a filed Class A count of 445,616,326 and a fully-exchanged total of 1,696,353,742, with the 10-Q stating that Interactive Brokers Group, Inc. owns "approximately 26.3% of the membership interests of IBG LLC", and priceToSalesRatioTTM of 13.79x rejected as wrong in both numerator and denominator, replaced with 22.2x on net revenues and a fully-exchanged equity value of $150.35B; every liquidity, leverage and turnover ratio rejected as a broker-balance-sheet artefact, including a current ratio of 17,791.6x and days-sales-outstanding of 4,077 days; enterpriseValueTTM and evToEBITDATTM rejected as inapplicable to a broker; est.ebitdaAvg/ebitAvg rejected for reporting EBIT above EBITDA in 2022 and 2023 and for a fixed 80.9% revenue ratio thereafter; epsdiluted NULL on every row, with diluted EPS taken from the earnings release; seg_prod and seg_geo recorded as BOTH EMPTY, with all revenue disaggregation taken from the 8-K; effectiveTaxRateTTM of 8.59% and taxBurdenTTM of 0.210 explained as partnership-structure artefacts rather than tax rates; earn_cal.revenueActual shown to differ from the filings by $11M for the March 2026 quarter and $42M for the June 2025 comparative; and all cash-flow multiples rejected on an operating-cash-flow-to-sales ratio of 151.6%. Where vendor and filing AGREED — worth recording: priceToEarningsRatioTTM of 35.03x is internally consistent and correct, being a Class A price divided by Class A earnings per share, and is the one headline multiple in this file that survives; netIncomePerShareTTM of $2.514 reconciles exactly to the four filed quarterly EPS figures; and the four-for-one forward split effective for holders of record 2025-06-16 is correctly applied throughout the payload, verified against the 10-Q's retroactive-adjustment statement — no split defect is present.
• Basis note: Interactive Brokers reports GAAP results and an adjusted measure, and for the June 2026 quarter the two are identical: "Reported and adjusted diluted earnings per share were both $0.69", with reported net revenues of $1.90 billion against $1.88 billion adjusted and reported pre-tax income of $1.46 billion against $1.44 billion adjusted. A GAAP-to-adjusted gap this small is rare and creditable, and it means no basis argument applies to this name. Analyst estimates are struck on diluted EPS attributable to common stockholders — the Class A slice — which is the same basis as the share price, so the price-to-earnings ratio is internally consistent.
• Estimate coverage: FIVE analysts on 2026 and 2027 net revenues; SIX and SEVEN on EPS; FOUR on 2028; TWO on 2029. This is the thinnest coverage in this batch on a $150 billion company, and it is consistent with the 19-analyst rating distribution and the unusually tight target range. The fair value is struck on the 2027 EPS distribution (7 analysts), whose high is only 9.7% above its low — a tightness we read as limited independent work rather than as agreement. The 2029 row is excluded from every conclusion.
• Peer note: the vendor peer set is a mixed financials basket — Bank of America ($446.4B), UBS ($175.2B), BBVA ($155.4B), Chubb ($134.4B), ICICI Bank ($108.5B), Brookfield ($99.7B), KKR ($97.2B), CME ($94.4B), Robinhood ($84.1B) and Intercontinental Exchange ($83.8B). Robinhood is the only genuine retail-brokerage comparable; CME and ICE are exchanges rather than brokers. No peer multiples are supplied in the file, so no peer-multiple comparison is drawn. One knowledge-base claim bears directly on the comparison — compound_and_friends (2026-06-23): "Prefer the brokers over the exchanges right now" — and is noted rather than used.
• Fair-value caveat: the $64 / $99 / $127 anchors are multiples of the 2027 consensus diluted EPS distribution — 20.1x the mean of $3.186 for the bear (cross-checked at 21.3x the low of $3.005), 31.1x the mean for the base (equivalently 27.2x the 2028 mean of $3.639), and 39.9x the mean for the bull (cross-checked at 34.9x the 2028 mean). Stated arithmetic, not a discounted cash flow. The base is sensitivity-disclosed: 28x the 2027 consensus gives $89 — essentially spot — and 34x gives $108, essentially the street. The base assumes the forward multiple broadly HOLDS on a like-for-like basis (27.8x on the 2027 estimate today against 27.2x on the 2028 estimate in twelve months), so the return is earnings growth rather than re-rating — which is the honest framing for a stock whose multiple has roughly doubled since our own knowledge base last called it a "reasonable ~15 PE" in January 2024.
• Timing: second-quarter 2026 results were released 2026-07-21, fourteen days before this dive, and the full release is in the archive. The next print is 2026-10-20, 77 days away — consensus $0.702 of diluted EPS on $1,932.9M of net revenues, which asks for only +1.9% sequentially on a quarter that grew 28.1% year on year. 2026-08-04 was an unremarkable +1.00% session on roughly $381 million of turnover; the only company-related item dated that day is a Form 4 recording a director's 25-share open-market purchase executed 2026-08-03.
• Accessibility note: no information in this dive is conveyed by colour. All emphasis is carried by bold text, table structure and explicit labelling.
• Not investment advice. Independent research, educational and informational only, never personalised. No recommendation to buy, sell or hold any security is made to any person.
• Version: 2026-08-04-full.