Night Watch's Roderick van Zuylen on Marex $MRX

Summary

• Roderick van Zuylen’s core thesis: Marex ($MRX) is a mispriced futures commission merchant trading at 7-8x this year’s earnings with a ~30% ROE, having “7x’d their earnings in the last 5 years” with 30%+ growth continuing post-IPO. The market treats it as a commoditized broker “similar to BGC or TP ICAP,” which “maybe 5 years ago would have been a fair characterization” — it has since become a tech-enabled clearing and prime-brokerage infrastructure business.
• Industry consolidation has left only a few readily buyable scaled players: 100 registered US FCMs 20 years ago, ~50 now, and below the JPMorgan/Goldman tier “there’s actually just three companies you can buy” — StoneX, Marex, and ADM’s unit, which is for sale. Banks are exiting under Basel III/IV capital rules while small players get priced out by tech and compliance costs, so acquisition bidding stays relatively uncompetitive and Marex keeps buying at “three or four times earnings.”
• The black-box tail risk is quantifiable and, in his assessment, survivable. FCMs sit between clearing house and client and historically lose $10-30M in no-fault blowups; the worst was 2020’s negative oil (ADM ~$200M, Interactive Brokers $100M+). Against $500M of excess capital, “losing 10 to 30 million… would be a bad quarter. The worst that has ever happened in this industry absent fault… would be a bad year.”
• “Your ROE ends up being a reflection of what you paid at your acquisitions.” Best specimen: TD Cowen’s prime brokerage, bought in December 2023 with ~$80M revenue, at $250M two years later with margins roughly doubled; the standard playbook — buy Middle East client relationships and capture “on day one… a 50% profit uplift” from lower CME/ICE clearing fees.
• The Ninky Research(?) short report of early October 2025 has “zero relevance going forward,” he argues. The core fake-profits allegation about unconsolidated Luxembourg SICAVs collapsed when management said “We consolidate those entities” — two unused funds with $2M of US equity — and Andrew said top executives bought shares the next day; he also recalled the company saying a global hedge fund had decided to move more business to Marex. Van Zuylen is quite confident, though acknowledges he could be wrong, that buyback authorization will be sought at the May AGM.
• Q4’s admission that volatility is past “Goldilocks” surprised the market and knocked the stock, but growth remained within the usual 10-20% algorithm. Van Zuylen reads customers shortening hedges as “deferred income,” and notes global futures volume growth stepped up from ~5%/year pre-COVID to ~12%; he thinks higher volatility was a major factor and floated improved structured-products cross-selling as another possibility. CME/ICE volumes were +12% YoY in January-February.
• The risk that keeps him up at night is Fed funds below 1.5%, not credit. A 1-point cut costs only ~5% of earnings — Marex shares margin interest with 60% of clients — but zero rates with 2.5-year hedges rolling off would take “a decent chunk out of earnings”; offsets are client balances growing from ~$12-14B to $20B and pricing power to claw back spreads. He expects a re-rate from 7-8x toward StoneX’s 12-15x, but “I’m not in it for the multiple expansion. That’s a nice bonus.”

Deep dive

1. Marex 101: a business the market still files under “commoditized”

• Van Zuylen’s plain-English version: Marex is a futures commission merchant — “the guys you call if you’re an airline or a hedge fund and you want to trade oil futures,” who then talk you into hedging FX and rate risk too. It IPO’d two years ago as a PE-sponsored roll-up; two private placements by the sponsor put pressure on the shares.
• The mispricing, as he frames it: the market sees “a commoditized financial services business… similar to BGC or TP ICAP,” fair five years ago but not now — it has grown into “a much higher quality financial infrastructure business” in a consolidated space: 100 FCMs 20 years ago, ~50 today, and below the bulge bracket effectively just StoneX, Marex, and ADM’s for-sale unit after StoneX bought RJ O’Brien.
• The demand backdrop: “we’re just trading more futures every year” — global futures volume growth stepped up from ~5%/year pre-COVID to double digits (~12%). He thinks higher volatility from Ukraine, COVID and other events was a major factor, while improved cross-selling into structured products was a possible additional factor.

2. Sizing the black box: history’s worst case is “a bad year”

• Andrew’s scar tissue frames the episode: he owned StoneX around the GCAP deal — “GCAP basically minted the entire acquisition price” in COVID volatility — but sold because it was “very black box-ish,” with write-offs like getting “stuck with a shipment of coal.” StoneX is up ~4x since.
• Van Zuylen’s tail-risk math: FCMs sit between clearing house and customer and eat the loss if a client misses margin. No-fault losses of $10-30M occurred repeatedly through history; the true outlier was negative oil in 2020 (ADM ~$200M, Interactive Brokers $100M+). Marex reports $500M of excess capital and does not want to run close to its regulatory limit because an investment-grade rating matters, so: “Losing 10 to 30 million… would be a bad quarter… The worst that has ever happened in this industry absent fault… would be a bad year.”

3. The ~30% ROE is a function of purchase price, not magic

• Why does Marex out-earn StoneX (~30% vs ~15% ROE) in “more or less the same business”? Partly leverage — StoneX’s CEO has “110% of his wealth invested” and runs it conservatively — but mostly deal prices: “Your ROE ends up being a reflection of what you paid at your acquisitions.” He also noted that StoneX has a distinct payments business and physical-trade exposure. ED&F Man came below book; recent market-making deals were at three or four times earnings.
• The showcase deal: TD Cowen’s prime brokerage, acquired in December 2023 shortly before the IPO at ~$80M revenue — $250M two years later, with margins roughly doubling on the fixed-cost base.
• The repeatable playbook: leverage clearing scale, buy a Middle East brokerage — really client relationships — and capture “on day one… a 50% profit uplift because they have lower clearing fees at CME and ICE.” Then sell the customer interest-rate futures alongside its oil hedge.

4. From strip-club voice broking to sticky clearing

• Van Zuylen is blunt about the legacy business: OTC voice brokerage where “you’re going to trade with the guy that took you out to a fancy restaurant yesterday or took you to the strip club,” brokers defect and go on garden leave — “It’s not a business you want to be in.”
• Clearing changed the quality: one to one-and-a-half years to onboard onto CME or ICE, half a year for clients to onboard onto Marex, heavy compliance and tech. Andrew’s switching-cost framing: change clearers and if “your trades don’t settle that day… you saved a little money and you blew the entire firm up.”
• Andrew’s cable-network pushback — shouldn’t rival bidders compete the day-one synergy away? The answer is the consolidation mechanics: Basel-constrained banks exiting above, compliance costs squeezing below, so bidding isn’t competitive. On RJ O’Brien he asked both sides and got “completely different answers”: StoneX says it wasn’t a competitive process; Marex asks why pay “six to nine times earnings post synergies” for a big platform “if you can buy three smaller ones for three or four times” — and it already had enough interest-rate risk.

5. Seven-to-eight times earnings, with the PE overhang partly cleared

• The valuation gap: 7-8x this year’s earnings versus StoneX at 12-15x; he thinks 20x is deserved but concedes the market may never fully trust the credit risk — “just because I think it’s worth 20 times, it doesn’t mean I’m ever going to get it.” His return math: “I’m not in it for the multiple expansion. That’s a nice bonus” — the 30%+ earnings growth (versus 10-20% guidance) does the work.
• The technicals: JRJ has been invested since 2009; the company’s 2022 UK IPO failed, and JRJ still owns 17%. Repeated private placements have forced the market to absorb stock, and shares dipped after the most recent placement two weeks before recording on block-placement rumors. Having that placement out of the way plus a longer public track record should help the multiple.
• Andrew’s PE-exit skepticism — “you’re kind of left holding the bag” — meets a results-based answer. Van Zuylen had exactly two IPO-time concerns: buying when PE sells, and financial-services roll-ups where “people are just loyal if you don’t pay them enough.” “That concern has been unfounded… They continue to execute, which in the end is what matters.”

6. Past Goldilocks, still compounding

• The surprise that hit the stock: on the Q4 call, with oil swinging from 60 to 100-120, management said current volatility “might be a bit too much to actually earn good money. We don’t want stress on our customers either” — clients may de-risk books rather than post margin, and an airline may hedge for months instead of years.
• Van Zuylen reads that as “deferred income” — the airline “is going to come back in 2 or 3 months and hedge more” — and stresses this is “not the kind of company that needs to make all of their earnings in 1 or 2 weeks of the year, like Flow Traders.” He said the business was still growing at least within its 10-20% algorithm; CME/ICE futures were +12% YoY in January-February, with energy and metals “way higher.”

7. The short report’s main allegations, as he sees it — and buybacks likely from May

• The Ninky Research(?) report: unconsolidated Luxembourg SICAVs allegedly booking “fake profits,” reduced auditor scope, and no real free cash flow. His triage — “it’s quite easy to make a report that looks scary” — but the key rebuttal came on the call: “We consolidate those entities.” If consolidated, “the entire argument is off the table”; the SICAVs were two unused funds with $2M of US equity, legacy of a roll-up with hundreds of legal entities.
• On other points, Van Zuylen said the semiannual cash-flow statements show free cash flow, though hedges can swing it from tailwind to headwind. Andrew relayed management’s explanation that the financing classification was typical for a financial firm and disclosed in footnote one. He also recalled management saying a regulator fine stayed with the seller because Marex bought assets rather than the legal entity, and that Deloitte’s scope changed because EY took over; he invited correction on the fine explanation.
• Andrew admired the response — no yelling, rebut the two scariest points, insider buying next day, and a “global hedge fund decided to move more business to us.” Van Zuylen’s personal frustration, kept as hedged: reports drop “5 minutes after the market opens… most of the time they’ve already covered part of their position. I’m not sure that it even happened” here — “but at the end of the day, let them say what they want to say and it keeps us sharp.”
• Buybacks: as a British/foreign entity listed in the US and reporting under IFRS, Marex needs shareholder approval, which prior PE owners did not want to provide. At the time, “everybody in the company had to chip in $1 million to buy back some shares.” His call: “I could be wrong. I’m quite confident that they will ask for buyback authorization at the next AGM, probably May.” And his real-life verification: launching his own fund 2.5 years ago, Marex’s TD Cowen unit “were the first ones to reach out” about prime brokerage, and he sees people moving business there for the swap lines.

8. The real risk: Fed funds below 1.5%

• The stated sensitivity is mild — a 1-point cut costs ~5% of earnings, because for 60% of clients Marex takes ~150bps of margin interest (versus Interactive Brokers’ 50bps), keeping all of it only on 40%. But if Fed funds hit zero and the 2.5-year rolling hedges expire, that’s “a decent chunk out of earnings.”
• The offsets: client balances grew from ~$12-14B to $20B in two years, and in a consolidated market both StoneX and Marex are “quite confident” they can reprice below 1.5% — “we gave you more money on the way up… how about you start paying slightly higher spreads.” Lower rates may also lift trading volumes in prime brokerage, less so clearing.
• Andrew’s closing pattern-match: IBKR and StoneX “really start taking off” in mid-to-late 2022 when rates rose — rates are the regime variable the market rewards — and with the Fed funds outlook shifting recently, “I’ve got good news for you on that risk.”