Interactive Brokers: Margin Masters - [Business Breakdowns, EP.216]

Summary

• Latitude’s Freddie Lait and Jacopo Di Nardo make account growth the central IBKR thesis. Headline per-account economics are relatively stable — ~200 trades/year at $3 average commission ($600) plus ~$800 of net interest margin — so with accounts up from ~1M to 3.5M in five years and growth inflecting from 25% toward 30–35%, “the real driver for this business over time has been and will be account growth,” with management targets of 10, 20 and eventually 80 million accounts.
• The moat is a “Costco model”: direct market access with no payment for order flow, much of the economics given back to customers. PFOF wholesalers’ profits “implicitly… come out of the customer’s execution price”; IBKR instead proves best execution, pays base rate less 50bps on cash balances over $10k, and charges margin rates roughly half of Fidelity or Schwab — funded by full automation that keeps headcount and opex “so much thinner than the others.”
• Founder Thomas Peterffy’s engineer-plus-risk-manager DNA is the culture: IBKR “doesn’t want to be in businesses that eventually cannot be automatized.” Its largest stated loss was ~1% of capital on the Swiss franc move; it refused duration risk at zero rates in 2021, stopped quoting long-dated options in 2007 before that market seized, and was expected to have ~$18B of excess capital by end-2025 — ~95% of total capital.
• Rate sensitivity is “probably one of the greatest misunderstandings in the business” — and helped create Latitude’s investment opportunity 18 months ago, when IBKR was trading at 12–13x delivered earnings. A 100bps global rate fall cuts NIM per account only ~10%, more than offset by 30%+ account growth, and margin loans move countercyclically to rates; contrast Schwab, which “almost needed effectively a wholesale rights issue” when its zero-paying deposits ran.
• More than half the business comes from B2B/non-retail cohorts. IBKR is the fifth-largest prime broker in the world from a standing start ~10 years ago, had just brought on HSBC globally as an introducing broker, and serves RIAs and prop traders — competitors are “folding and using this business instead of trying to compete with it,” renting the tech the way institutions outsource custody to JPMorgan or Northern Trust.
• M&A math structurally fails, so growth is primarily organic — and there’s an unpulled lever. Any acquired book’s revenue and profits “would collapse” once repriced to IBKR’s spreads; meanwhile Peterffy says “we can grow at 30% with no advertising. We can probably grow at 40, 50% if we do more targeted advertising” — a deliberate reserve.
• The inversion frames the upside; ownership and regulation frame the risks. At ~$1,500 revenue per account and roughly 75% margins, 10–20M accounts implies $15–30B of revenue — “the question one really needs to ask is who’s going to stop that happening?” Offsets: insiders own ~80% (Peterffy ~75%), making large capital returns not necessarily doable until the float grows; the current capital return is a usually targeted 0.5–1% dividend yield. Jacopo also flags gambling-adjacent retail activity: the CFTC initially blocked ForecastEx as too akin to sports betting, creating potential regulatory risk.

Deep dive

1. Lowest cost to serve, highest quality output — account growth is the key driver

• Freddie Lait’s setup: through 50 years of investing in technology and automation, IBKR built “by far and away the lowest cost model, the lowest cost to serve with the highest quality output,” a flywheel compounding “without advertising, without marketing” — and now doubling up through other channels: introducing brokers, hedge funds, RIAs, prop traders and genuine international reach.
• The unit economics are simple to model: 3.5M accounts × ~200 trades/year × ~$3 average commission ≈ $600 per account, or ~$2B in commissions; net interest margin — a securities-lending and margin book of ~11% of client assets financed by customer deposits — adds ~$800 per account, ~$3B. 2024 P&L: ~$1.7B commissions, ~$2B total non-interest income, ~$3B NII, ~$5B revenue.
• Jacopo Di Nardo on who uses it: the most automated of all brokers — even liquidating accounts that breach margin limits “is done automatically by a bot” — attracting experienced traders with $200k to a couple million per account. Robinhood or eToro accounts might average $5–10k, while Schwab/Fidelity accounts are used much less frequently, with 401(k)s one U.S. example.

2. Peterffy’s imprint: automate everything, manage risk

• The founder arrived from Hungary with nothing, bought a seat on the Chicago Board Options Exchange as Black-Scholes formulas were emerging, and built Timber Hill as an options market maker on one principle: “he wanted to automatize everything that could be automatized. And this still permeates the culture… Interactive doesn’t want to be in businesses that eventually cannot be automatized.”
• The risk record is the proof: the largest stated loss was ~1% of capital when the Swiss National Bank let the franc appreciate freely; at zero rates in 2021, while other brokers took duration risk to juice NII, “the company refrained”; in 2007 it stopped making prices in long-dated options before that market became entirely illiquid.
• He’s also “a very hard-nosed and smart businessman”: when post-GFC regulation and the market environment meant options market making would no longer earn excess ROE, he shut down most of that business by 2014 and focused entirely on online brokerage.

3. The moat: no PFOF, honest spreads, a fortress balance sheet

• On payment for order flow — “pretty much just a US phenomenon” — Lait’s framing is a Costco analogy: IBKR is “cutting out the wholesaler,” linked directly to essentially every exchange in the world, because PFOF buyers’ profits “have to come out of the customer’s execution price.” IBKR forgoes the revenue, proves best execution, and still undercuts on commissions.
• The deposit arrangement is “just far more honest”: base rate less 50bps on cash balances over $10k, representing roughly two-thirds of customer cash. Contrast Schwab, which paid nothing, watched customers flee to money market funds when rates rose, and “almost needed effectively a wholesale rights issue.”
• Di Nardo estimated ~$18B of excess capital by end-2025, ~95% of the total. Automated liquidation — no Archegos-style human phone calls — is what lets IBKR price margin loans at about half of Fidelity or Schwab while offering slightly more leverage at lower actual risk.

4. More than half the business is B2B — and the NIM fear is the misunderstanding

• More than 50% of the business, and much of the growth, comes from three non-retail cohorts: RIAs, prop traders and hedge funds — IBKR is now “the fifth largest prime broker in the world… larger than many of the big banks” — plus introducing brokers like HSBC, which IBKR had just brought on globally and planned to connect to its underlying trading capability, the way institutions rent custody from JPMorgan or Northern Trust.
• The mix math: trades per account drift down 5–10% per year and commissions perhaps -5%, “but if account growth is 35%… that will dwarf any reinvestment back in price.” On rates: a 100bps global fall (IBKR is ~40% US) cuts NIM per account only ~10% — “that’s why we focus so obsessively on… the potential for account growth.”
• Di Nardo adds the offset: if rates returned to 2019–21 levels, margin loans per account would likely be materially higher — partially offsetting an NIM fall.
• The runway: Lait estimated Schwab at ~25M mostly-US accounts and thought Fidelity was around that level; IBKR’s 3.5M is a fraction, with 20M globally “a sort of first stop” and 80M the long-run number (already ~1% population penetration in some countries). Robinhood/eToro are feeders: customers may move to IBKR once balances hit $100–200k. Lait’s Ryanair parallel — O’Leary’s “I didn’t realize looking after customers would make me so much profit” — maps to IBKR’s new hedge-fund “white glove” service and app relaunch.

5. Valuation by inversion; ownership, regulation and lessons

• Valuation is genuinely awkward. Jacopo said ROE might be 15–20% including ~$19B excess capital and “10 to 15 times as much” excluding it; they also use P/E on normalized earnings. Lait inverts: ~$1,500 revenue per account, roughly 75% margins, and 10–20M accounts implies $15–30B revenue. “The question one really needs to ask is who’s going to stop that happening?” Peterffy claims 30% growth with no advertising, “probably 40, 50%” with it — a lever unpulled.
• Capital returns are structurally constrained: a dividend targeting 0.5–1% of market cap is currently the only return, while insiders own ~80% (Peterffy ~75%), making large special dividends or buybacks not necessarily doable until the float grows. Di Nardo’s hedge: Peterffy “is not immortal” and is approaching or above 80; the excess capital earns returns in the meantime — the Berkshire parallel.
• Jacopo says investors should take comfort from the liquidity profile: most assets are ~30 days in duration. Recent drawdowns produced “de minimis” margin damage, but he also flags potential regulatory scrutiny of gambling-adjacent retail flow: the CFTC initially blocked ForecastEx, a platform for yes/no views on outcomes such as inflation or payrolls, as “too akin to sports betting,” which “tells you there is a component in financial markets today… more similar to gambling.”
• The closing lessons: low-cost/high-service models (Ryanair, Costco) are “so hard to compete with”; the culture can’t be installed — “I don’t think it’s something that you can come through with an MBA… it’s got to be something that’s really in you”; and high-quality cyclicals get mispriced on cyclicality fear — IBKR traded at 12–13x delivered earnings 18 months ago. “The great ones are worth waiting for.”