MarketAxess Holdings Inc. (MKTX) - Deep Dive Research Report

Prepared 2026-06-12. Financial-year basis: calendar year. Most recent reporting period: Q1 2026 (quarter ended 31 March 2026), reported 6 May 2026.

1. What the company does

MarketAxess runs the plumbing that lets big bond investors and the banks that trade with them buy and sell corporate bonds on a screen instead of over the phone. When a pension fund or an asset manager like PIMCO wants to sell $5 million of an Apple bond, they used to call three or four dealers, ask each for a price, and pick the best one. That process was slow, opaque, and left the buyer guessing whether they got a fair price. MarketAxess turned that phone call into an electronic request that fans out to dozens or hundreds of potential counterparties at once, collects competing prices in seconds, and prints the trade with an audit trail. The company gets paid a small commission on each bond that changes hands on its system.

The business was born inside a bank. In 1999, Richard McVey, then running JP Morgan's institutional e-commerce effort, pitched the idea through the bank's "LabMorgan" incubator. In April 2000 it spun out as an independent company with two employees, roughly $24 million of capital, and windowless offices that JP Morgan let McVey use rent-free. At the time there was essentially no electronic trading in corporate bonds; the whole market ran on phone calls, faxes, and Bloomberg messages. MarketAxess started with the most basic electronic protocol, the multi-dealer "request for quote" (RFQ), where a client asks several dealers for a price at once. It went public in 2004.

The pivotal decision in the company's history came around 2012-2013 with the launch of Open Trading. The 2008 crisis and the Volcker Rule had forced banks to shrink their bond inventories, which meant the old model (where a dealer used its own balance sheet to take the other side of every client trade) was breaking down. Open Trading flipped the model: it created an anonymous "all-to-all" marketplace where any participant, including another asset manager or a hedge fund, could be the counterparty, not just a dealer. This turned MarketAxess from a tool that connected clients to dealers into a genuine network where liquidity could come from anywhere. That network is the heart of the company today.

"Block trading represents the next step function to the growth of electronic trading." - CEO Christopher Concannon, Q3 2025 call (7 November 2025). The line captures how management thinks: the company has spent 25 years moving ever-larger and ever-harder slices of the bond market onto a screen, one protocol at a time.

The technical difficulty is not the software, it is the network and the data. A bond-trading venue is only useful if enough buyers and sellers are already there, which is a classic chicken-and-egg liquidity problem that took more than a decade to solve. Layered on top is CP+, MarketAxess's AI-driven composite pricing engine, which estimates a fair price for hundreds of thousands of individual bonds that may not have traded all day. Unlike equities, where one Apple share is identical to the next, a single company can have dozens of different bonds outstanding, each with a different maturity and coupon, most of which trade rarely. Pricing that universe accurately, in real time, off a proprietary dataset of actual executed trades, is the moat that a new entrant cannot simply code around.

A concrete walk-through: a credit portfolio manager needs to sell 40 different bonds at once to fund redemptions. Rather than work 40 separate orders, she uses MarketAxess portfolio trading to package all 40 into one list, sends it to competing dealers who each price the whole basket, and executes the entire thing in a single click against the best bid. CP+ tells her what each line should be worth, Open Trading widens the pool of who can price it, and the post-trade system reports it to regulators automatically. MarketAxess collects a commission on the notional traded. That is the business.

2. Business segments

MarketAxess is effectively a single-business company (electronic fixed-income trading), but it reports revenue across a few distinct lines, and the more useful way to understand it is by revenue type and by product channel. I will treat the three revenue lines as segments, then describe the product channels in Section 3.

Commissions (the engine - roughly 85%+ of revenue)

This is the core. MarketAxess charges a commission on every bond traded on the platform, structured as a mix of variable per-trade fees plus fixed monthly distribution fees that dealers pay for connectivity. Within commissions, credit (corporate bonds: US high-grade, US high-yield, emerging markets, eurobonds, municipals) is overwhelmingly dominant, contributing the large majority of commission revenue, while rates (US Treasuries and agency bonds) is a much smaller, lower-fee-per-trade business. In Q1 2026, credit commission revenue grew 9% and rates commission revenue grew 29% off a small base.

The core capability here is the two-sided liquidity network plus the pricing data that makes anonymous trading safe. This is what took 25 years to build and is the hardest thing to replicate. It is the margin engine and the cash cow simultaneously: commissions scale with trading volume at very high incremental margin because the platform is already built.

Information Services (roughly 7% of revenue)

MarketAxess sells market data: real-time and historical pricing feeds, the CP+ composite price product, axe data (what dealers want to buy or sell), and analytics. Because the company sees a huge share of US credit trades, it sits on a uniquely rich dataset, and it monetises that exhaust by selling it back to the market as subscriptions. This is a recurring, sticky, subscription revenue line that management guided to "mid-single digit" growth for 2026 (Q4 2025 call). It exists as a separate line because the economics are different: subscription, not transactional, and it deepens the moat (the more trades flow through, the better the data, the more valuable the feed).

Post-Trade Services (roughly 5% of revenue)

This is regulatory reporting and trade matching - the back-office plumbing that happens after a trade is agreed. MarketAxess built this through acquisitions: Xtrakter/Trax in 2013 and Deutsche Börse's Regulatory Reporting Hub in 2020-2021, giving it a European regulatory-reporting franchise (MiFID II, EMIR, SFTR reporting). It exists separately because it is a different regulatory product with a different (largely European) customer base, and it is somewhat strategic: it keeps MarketAxess embedded in client workflows beyond execution.

3. Products and business detail

MarketAxess is best understood as one liquidity network accessed through many trading protocols. The protocols are the products, and each one is an attempt to digitise a slice of the bond market that previously stayed on the phone.

The product channels (by asset class):

• US High-Grade (investment-grade) credit - the original franchise and largest single product.
• US High-Yield credit - junk bonds; harder to trade electronically, larger trade sizes.
• Emerging Markets - hard-currency and increasingly local-currency bonds; a major international growth driver, with record volume up 30% in Q1 2026 and 1,547 active client firms.
• Eurobonds - European corporate credit; block volume up 45% in Q1 2026.
• Municipals - US muni bonds, entered via the 2021 MuniBrokers acquisition.
• Rates - US Treasuries and agencies, built out via the 2019 LiquidityEdge acquisition; lower fee per trade but high volume.

The protocols (how trades happen):

• RFQ (Request for Quote) - the original protocol; client asks multiple dealers for a price.
• Open Trading - the anonymous all-to-all marketplace; any participant can provide liquidity. Penetration reached 47% of US credit volume in Q1 2026, the highest since 2023. This is the crown jewel and the source of price improvement that keeps clients loyal.
• Portfolio Trading (PT) - execute a whole basket of bonds in one transaction. Global PT ADV rose 51% to a record $1.9 billion in Q1 2026; US high-yield PT ADV jumped roughly 78%. This is the protocol where MarketAxess has historically been weakest versus Tradeweb and is now investing hard to catch up.
• Block Trading - large institutional-size trades, the hardest to move electronically because of market-impact risk. Block ADV grew 35% to a record $7 billion in Q1 2026, and dealer algos captured 30% of US high-grade block trades in March 2026. Management calls block "the next step function" of electronic growth.
• Mid-X / dealer-initiated - a dealer-to-dealer midpoint matching session launched in US credit in September 2025; it matched $1.3 billion in its first ten days and generated $7 billion of volume in 2025 (up 383%). This opens an entirely new dealer-to-dealer revenue pool.
• Automation Suite (Auto-X, Adaptive Auto-X) - algorithmic trading that lets clients automate small and mid-size orders by rule. Record automation volume of $144 billion in Q1 2026; Adaptive AutoX exceeded $8 billion in a single quarter (Q4 2025). Automation grew at roughly a 29% CAGR on trade count.
• X-Pro - the next-generation trading front-end (the screen traders actually use), bundling pricing, analytics, and all protocols in one workflow.

Data and technology products: CP+ (AI composite pricing, now across all products), the broader Information Services feeds, and execution-analytics tooling acquired with Pragma in 2024. A new DirectBooks partnership is launching in summer 2026 to give clients real-time access to the new-issue calendar and streamlined post-issuance block trading, addressing the primary (new-issue) market where MarketAxess has historically had no presence.

Geographies: headquartered in New York, with major operations in London, Amsterdam, Singapore, Hong Kong, and across Latin America. International product growth (EM, eurobonds, local markets) is now the fastest-growing part of the company, with revenue outside US credit up 20% in Q1 2026.

Acquisition milestones that reshaped the business: LiquidityEdge (2019, rates), Regulatory Reporting Hub (2020-2021, European post-trade), MuniBrokers (2021, municipals), Pragma (2024, algorithmic execution and analytics), and RFQ-hub (2025, multi-asset/derivatives RFQ, added to international revenue in 2025).

4. Customers

The customer base splits into two interdependent groups: the buy-side (asset managers, hedge funds, insurance companies, pension funds) who initiate most trades, and the sell-side dealers (the big banks) who provide a large part of the liquidity and pay connectivity fees. A healthy network needs both.

On the buy-side, the buying decision is made by a firm's head of trading and its portfolio managers, with input from compliance. The criteria are concrete: best execution (did I get the best available price?), the depth of available liquidity (can I actually fill a large or illiquid order?), the quality of pricing data (CP+), and the breadth of workflow tools (can I automate the small stuff and use one screen for everything?). Regulatory best-execution obligations actually push flow toward MarketAxess, because an electronic venue produces the auditable proof of competitive pricing that a phone call cannot. In Q1 2026 the company reported 1,547 active client firms and 3,410 active traders in emerging markets alone.

The reason clients choose MarketAxess is the network itself: Open Trading routinely delivers measurable price improvement (clients save money versus the dealer-only price) precisely because the pool of counterparties is wider. That price improvement is quantified and reported back to clients, which is a powerful, data-backed retention argument that competitors with thinner networks struggle to match.

Switching costs are real but not absolute. The platform is deeply embedded in trader workflows (X-Pro, automation rules, API connections to order-management systems), and a client's automation setup represents months of configuration. But the binding constraint is liquidity, not contracts: traders go where the liquidity is, and most large institutions are connected to MarketAxess, Tradeweb, Bloomberg, and Trumid simultaneously and route each order to whichever shows the best price. This multi-homing is the central competitive reality of the business.

Concentration is low on the buy-side (thousands of client firms), which makes revenue resilient to any single client leaving. On the dealer side there is more concentration among the large global banks, but their connectivity fees are a smaller part of revenue than buy-side transaction commissions. Contract structure is predominantly transactional: MarketAxess earns commissions on volume, supplemented by recurring distribution fees from dealers and subscription fees from Information Services. This makes revenue partly volume-dependent (and therefore sensitive to market activity and volatility) but with a growing recurring base from data and post-trade.

5. Competitive landscape

Electronic credit trading is an oligopoly with one structural feature that shapes everything: clients multi-home. Because every large institution is connected to all the major venues, market share is won order-by-order on price and liquidity, not by locking customers in. This keeps competition perpetually intense even though there are only a handful of serious players.

The real competitors:

• Tradeweb (TW) - the primary rival and the most important competitor to watch. Historically the leader in rates and in portfolio trading (over two-thirds of e-portfolio-trading volume in 2024), and now pushing aggressively into US credit, where it has been gaining share faster than the market grows. Tradeweb is the one competitor capable of attacking MarketAxess's core.
• Trumid - a private, venture-backed US credit upstart built around anonymous all-to-all trading, directly attacking MarketAxess's Open Trading franchise. Small but growing fast (the largest percentage volume increase in US credit e-trading in 2024).
• Bloomberg - private, terminal-based; strong in RFQ, especially in eurobonds and emerging markets where the terminal is ubiquitous. Its distribution (every trader has a terminal) is a structural advantage.
• Intercontinental Exchange (ICE) - runs ICE Bonds (the former BondPoint and TMC platforms); a credible but secondary credit venue, and a giant overall.
• CME Group, BGC Group (Fenics), Euronext (MTS) - relevant mainly in rates, less so in credit.

Where MarketAxess wins: US credit market share (roughly 37% of US corporate-bond e-trading versus Tradeweb's ~34%, Trumid ~13%, Bloomberg ~11% per Coalition Greenwich 2024-25), the depth of Open Trading liquidity, the CP+ data advantage built from the largest pool of executed credit trades, and a 25-year head start on the network.

Where MarketAxess is exposed: portfolio trading, where Tradeweb has been the clear leader and MarketAxess is the challenger (it gained ~100bps of US PT share in Q1 2026 but is catching up from behind); and the broad multi-homing dynamic that means no win is permanent. The structural shift to watch is the migration from RFQ toward portfolio trading and block trading, both protocols where the competitive order is less settled than in plain RFQ.

Barriers to entry are high for a generalist but lower for a focused attacker. Building a from-scratch generalist credit venue with no liquidity and no pricing data is close to impossible (Trumid has needed years and heavy funding just to reach low-double-digit share). But a well-capitalised incumbent like Tradeweb, with an existing rates network and buy-side relationships, can credibly extend into credit, which is exactly what is happening.

6. Industry

Demand for MarketAxess's services is driven by two forces: the secular shift of bond trading from voice to electronic, and the overall level of bond-market activity (issuance and trading volume, which rises with volatility and with the size of the outstanding bond market).

The US corporate bond market is roughly $11-12 trillion in outstanding notional, and the broader fixed-income universe (rates, EM, eurobonds, munis) is many times larger. The key statistic is electronification: nearly half of US investment-grade corporate bond volume now trades electronically, up from essentially zero when MarketAxess started in 2000, and high-yield is following behind. Coalition Greenwich has noted that e-trading growth is expected to outpace overall bond-market growth, which means the addressable electronic pool expands even in a flat market. Within that, the still-voice-traded segments (large blocks, illiquid names, new issues) are the frontier, and they are exactly what block trading, Mid-X, and the DirectBooks partnership are designed to capture.

A second structural tailwind is indexation. As fixed-income ETFs grow (management cited ETF assets potentially reaching $5 trillion), more trading becomes systematic and rules-based, which favours electronic, automated execution and basket (portfolio) trades. MarketAxess launched a closing-auction protocol in late 2025 specifically to serve this indexed flow. Systematic hedge funds and large quantitative managers are an increasingly important liquidity source on Open Trading.

Where MarketAxess sits in the supply chain: it is the venue layer between the buy-side and the dealers, plus the data layer that prices the market and the post-trade layer that reports it. It does not take principal risk or hold inventory; it is an infrastructure and data toll-taker.

Regulation is broadly a tailwind. Best-execution requirements push flow onto auditable electronic venues; post-crisis rules that shrank dealer balance sheets created the need for all-to-all liquidity in the first place; and European reporting mandates (MiFID II, EMIR, SFTR) created the post-trade business. The main regulatory risk is the opposite direction: any rule that fragments liquidity or caps electronic-venue economics.

Cyclicality: the business is moderately cyclical but in an unusual way. Revenue is volume-driven, and volumes rise with volatility and with credit issuance. Periods of low volatility and tight credit spreads (as in Q3 2025, when "a return to lower levels of volatility and tighter credit spreads ... dampened market share and revenue growth in U.S. credit") compress activity and fee capture, while volatile periods boost volumes. The secular electronification trend is the offset that keeps the long-run trajectory upward through the cycle.

7. Growth triggers

All points below are drawn directly from the five earnings calls. Each is attributed.

• Block trading scaling into the largest untapped pool. Block ADV grew 35% to a record $7 billion in Q1 2026, with dealer algos taking 30% of US high-grade block trades in March 2026. Management frames block as the next leg of electronic growth. (Q1 2026 call, 6 May 2026; repeated theme from Q3 2025, 7 Nov 2025, and Q4 2025, 6 Feb 2026.)

"Block trading represents the next step function to the growth of electronic trading." - Concannon, Q3 2025 call (7 Nov 2025). Repeated across all five calls, making it the company's central growth narrative.

• Portfolio trading share gains. Global PT ADV rose 51% to a record $1.9 billion in Q1 2026, with US credit PT share up ~100bps and US high-yield PT ADV up ~78%. (Q1 2026 call, 6 May 2026; PT share gains cited in every call - up 370bps in Q2 2025, 210bps YTD in Q3 2025, 270bps for full-year 2025.)

• Mid-X / dealer-initiated trading opening a new revenue pool. Launched September 2025; $1.3 billion matched in the first ten days, $7 billion total in 2025 (up 383%), and management plans to expand beyond the current single daily dealer session. (Q4 2025 call, 6 Feb 2026; Q3 2025 call, 7 Nov 2025.)

• DirectBooks partnership for new-issue trading, launching summer 2026. Gives clients real-time access to the new-issue calendar and streamlined post-issuance block trading, addressing the primary market MarketAxess has not previously served. (Q1 2026 call, 6 May 2026 - new.)

"We are addressing that new issue market with launch of our direct books partnership." - Concannon, Q1 2026 call (6 May 2026).

• Automation and AI-driven algos. Record automation volume of $144 billion in Q1 2026; Adaptive AutoX exceeded $8 billion in Q4 2025; CP+ AI pricing now across all products. Automation growing at roughly a 29% CAGR. (Q1 2026 call, 6 May 2026; Q4 2025 call, 6 Feb 2026.)

• Emerging markets and eurobonds as the international growth engine. EM trading volume up 30% to record levels in Q1 2026 (local-markets revenue up 56%); record EM ADV of $5 billion in January 2026, up 50% YoY; revenue outside US credit up 20% in Q1 2026. (Q1 2026 call, 6 May 2026; Q4 2025 call, 6 Feb 2026.)

• Indexation / closing-auction protocol. A new closing-auction protocol launched in late 2025 to serve the growing indexed/ETF fixed-income market, with ETF assets potentially reaching $5 trillion. (Q3 2025 call, 7 Nov 2025 - new.)

• New initiatives now driving the topline. "Momentum continued to build with our new initiatives and generated approximately 50% of total incremental revenue in the quarter." - Concannon, Q1 2026 call (6 May 2026). This is the clearest signal that the newer protocols (block, PT, automation, EM) are now material, not experimental.

8. Key risks

Tradeweb taking US credit share. This is the central risk to the franchise. Tradeweb already leads in portfolio trading and rates and has been extending into US credit faster than the market grows. Because clients multi-home, share can shift order-by-order without any contract change. The mechanism: if Tradeweb's credit liquidity reaches a tipping point, traders route more credit flow there, MarketAxess's volume-based commissions erode, and its data advantage (which depends on seeing the most flow) slowly degrades. This is high-probability as a continued grind, lower-probability as a sudden collapse, but it is the risk that most directly threatens the core.

Low-volatility, tight-spread environments compress revenue. The business is volume-driven, and volumes fall when markets are calm. Management said this plainly:

"A return to lower levels of volatility and tighter credit spreads in the third quarter, combined with historic levels of new issuance in September, dampened market share and revenue growth in U.S. credit." - paraphrasing the Q3 2025 disclosure (7 Nov 2025).

This is a high-probability, moderate-drag risk: it recurs whenever credit markets are quiet, and Q3 2025 showed total revenue growth slowing to just 1%.

The US high-grade core is maturing. Electronification of investment-grade is already approaching half the market, so the easy share gains in the original franchise are largely behind the company. Future growth depends on newer, less-proven protocols (block, PT, Mid-X, new issue) and international products. If those ramp slower than hoped, the company's growth rate de-rates toward the maturity of its legacy business. Q3 2025's flat revenue is a preview of what a maturing core looks like when the new initiatives have a soft quarter.

Disruption from a focused all-to-all attacker. Trumid is purpose-built to attack Open Trading, MarketAxess's most valuable franchise, with the largest percentage volume growth in US credit e-trading. If an anonymous-network competitor reaches critical liquidity in specific segments, it can peel off exactly the high-value flow MarketAxess depends on. Lower-probability today given share is still ~13%, but structurally the most direct threat to the crown jewel.

Execution risk on the build-out. The growth thesis rests on the company successfully shipping and scaling a long list of new protocols simultaneously (block, Mid-X, closing auctions, DirectBooks, X-Pro, AI refactoring of legacy code). Management is spending into this, with 2026 operating-expense growth guided to ~8% ($530-545 million) and capex of $65-75 million, 80% on software. If the new initiatives underdeliver while costs keep rising, operating margin compresses. Moderate probability, and directly tied to the heavy investment cadence management has committed to.

Pricing and fee pressure. Because clients multi-home and best-execution rules reward price competition, there is persistent downward pressure on commission rates. A price war initiated by a deep-pocketed rival (Tradeweb, or ICE) to buy credit share could compress fee capture even if volumes hold.

9. Walk the talk

The five calls used for this assessment:

• Q1 2025 - 7 May 2025
• Q2 2025 - 6 August 2025
• Q3 2025 - 7 November 2025
• Q4 2025 - 6 February 2026
• Q1 2026 - 6 May 2026 (within 90 days of today)

The throughline across these five calls is a management team that has repeatedly framed the company's future around a specific set of new protocols (block trading, portfolio trading, automation, Mid-X, international) and has, for the most part, delivered measurable progress on each, even through a soft patch in the middle of the period.

Starting with Q1 2025 (May 2025), Concannon set the tone by pointing to record daily volumes across most product areas, including record portfolio trading and record block trading ADV in EM and eurobonds, and positioned block and PT as the engines of future growth. The promise was momentum in the newer protocols.

By Q2 2025 (August 2025), that promise was being delivered on the protocol metrics: record block trading volume across US credit, EM, and eurobonds (YTD block up over 20%, ~$8 billion cumulative since launch), record PT ADV, and US high-grade PT share "over 19% in the quarter, up 370 basis points over the prior year." Management also pre-announced the September Mid-X launch. Notably, GAAP EPS of $1.91 slightly missed consensus ($1.94), an early sign that headline financials were not keeping pace with the operational records, even as the protocol story tracked.

Q3 2025 (November 2025) is the most important credibility test in the window, because this is where the talk and the financial results diverged. Total revenue grew just 1%, and management openly attributed it to low volatility, tight spreads, and heavy new issuance dampening US credit. Critically, they did not spin it: they disclosed the headwind plainly and pointed to the offsets that were working - Mid-X launched on schedule in September (matching $1.3 billion in ten days), PT share up 210bps YTD, automation growing ~29%, and Open Trading hitting 34-39% of October volume, the highest since the 2023 regional-bank crisis. Concannon's framing was candid:

"We are confident that we can execute faster to generate higher levels of growth." - Q3 2025 call (7 Nov 2025), said in the context of acknowledging US credit revenue challenges rather than denying them.

That candor matters: management owned a weak quarter and attributed it to the market environment, not to a broken strategy, and the protocol metrics they had promised in prior quarters kept compounding through the soft patch.

By Q4 2025 (February 2026), the full-year picture validated the multi-quarter narrative: record full-year revenue ($846 million) with 10% non-US-credit growth, block up 24% for the year (January 2026 block surged 56% YoY to one-third of credit ADV), PT ADV up 48% with US credit PT share up 270bps, and Mid-X generating $7 billion in 2025 (up 383%). Management also set clear, specific 2026 guidance: opex of $530-545 million (~8% growth), tax rate 24-26%, capex $65-75 million, and mid-single-digit Information Services growth. These are falsifiable numbers, which is itself a credibility marker.

Q1 2026 (May 2026) then delivered against that newly set bar: total revenue up 12% to a record (a clear re-acceleration from the 1% of Q3 2025), 20% non-US-credit growth, block ADV up 35%, global PT ADV up 51%, EM volume up 30%, operating margin up ~200bps to 44%, and EPS up 20%. Most tellingly, Concannon quantified the strategy's payoff: new initiatives "generated approximately 50% of total incremental revenue in the quarter." The re-acceleration from a 1% quarter to a 12% quarter within two reporting periods is the strongest evidence that the soft Q3 was a market-environment dip rather than a strategy failure.

Assessment: this is management that does broadly what it says. The protocol-level promises (block, PT, Mid-X, automation, international) have been delivered with specific, verifiable metrics across five quarters, including through a genuinely weak revenue quarter that they disclosed honestly rather than obscuring. The fair criticism is that headline financial results have at times lagged the operational records (the Q2 2025 EPS miss, the Q3 2025 1% revenue growth), which reflects the volume-and-volatility dependence of the model more than any credibility gap. They guide specifically and then report against those specifics. On the spectrum from overpromise to underpromise, this team sits closer to "consistently accurate and candid," with a mild tendency to let strong operational metrics carry the narrative through softer financial quarters.

10. Shareholder friendliness index

Dividends. MarketAxess pays a growing quarterly dividend and has raised it steadily: $0.72 per quarter through 2023, $0.74 in 2024, $0.76 in 2025, and $0.78 declared for 2026 (paid 3 June 2026). That is roughly a 3% annual increase each year - a slow, reliable grower rather than a high-yield payer, consistent with a profitable, cash-generative business that prefers to retain optionality for buybacks and tuck-in acquisitions. There is no sign of an unusual payout pattern (no cut, no special dividend), and the payout is comfortably covered by free cash flow ($316 million on a trailing-12-month basis as of Q1 2026, $347 million for full-year 2025).

Buybacks and dilution. Buybacks have accelerated meaningfully. Last ~90 days (MoatMap window, since 14 March 2026): MoatMap recorded zero buyback transactions in this trailing-90-day window, but the company's own disclosures show it returned $60 million via repurchases in Q1 2026 around the completion of its accelerated share repurchase, with $205 million of authorization remaining as of 30 April 2026 (so the absence in the MoatMap block reflects feed coverage, not a halt in repurchasing). Older programmes (from annual reports and 8-K filings): in full-year 2024 the company repurchased 341,477 shares for $75 million, and on 6 August 2024 the board added a new $200 million authorization on top of the $50 million then remaining. The standout event was a $300 million accelerated share repurchase completed in early February 2026, which retired roughly 1.7 million shares, contributing to a ~6% reduction in diluted share count year-over-year by Q1 2026. Across the three-year window the share count has moved from roughly flat (modest option dilution offset by buybacks in 2023-2024) to clearly shrinking in 2025-2026 as repurchase intensity stepped up. Total capital returned to shareholders in 2025 was $474 million via buybacks and dividends.

Verdict: Returns Capital - a steadily rising dividend plus an accelerated buyback that is now actively shrinking the share count, funded comfortably by free cash flow.

11. Insider activities

Source: the MoatMap cross-market disclosure database (US venue; SEC Form 4 spine), data current as of 2026-06-11, cross-checked against the most recent filing window. The block is treated as authoritative for filings older than ~2 weeks; the most recent transaction (3 June 2026) is within the cross-check window and is consistent with routine compensation-related activity.

Recent transactions (most recent first):

Buys - read the signal. There were no open-market purchases by any insider in the last 12 months. The CEO and CFO "Other" transactions are equity-award or deemed-acquisition entries (dividend-equivalent or RSU-related), not conviction buys, and the CTO's 218-share line at $0 is a standard new-executive equity grant. There is no cluster buying and no first-time CEO/CFO open-market purchase to flag. The absence of any open-market buying is the relevant fact here: insiders did not put fresh personal capital into the stock during the window.

Sells - work out the why. The only open-market sale was a token 100-share disposal by General Counsel Scott Pintoff on 10 April 2026 (~US$17,200), which is immaterial relative to a senior executive's compensation and holdings - the kind of routine, sub-threshold housekeeping that carries no signal. No reason was disclosed in the filing footnote, and none is inferable beyond ordinary personal liquidity; given the trivial size, "reason not disclosed" is the honest read, and it does not matter. Notably, there were no large 10b5-1 program sales, no founder diversification blocks, and no sponsor exits in the window.

Net assessment. Insider activity over the last 12 months is best described as neutral and immaterial. The picture is dominated by routine equity grants and dividend-equivalent entries ("Other"), with a single trivial 100-share sale and zero open-market purchases. Activity is not concentrated in any one insider in a signal-bearing way, and nothing has changed recently that would shift the read in either direction. There is no bullish cluster-buying signal and no red-flag selling - just the ordinary compensation mechanics of a profitable company. Plain-language read: neutral. (Insider conviction is simply absent from the tape here; the investment case has to rest on the operating and capital-return evidence in Sections 7-10, not on insider buying.)

12. Scenarios

Bull case. MarketAxess successfully converts itself from a maturing US investment-grade RFQ business into the dominant venue for the next generation of bond-trading protocols. Block trading, today the largest still-voice-traded pool, keeps electronifying and MarketAxess's dealer-algo and Mid-X tools make it the natural home for it; portfolio trading share keeps climbing until the company is genuinely competitive with Tradeweb in baskets rather than a distant second; the DirectBooks partnership cracks open the new-issue market that the company never previously touched; and international (EM, eurobonds, local markets) compounds at double digits, diversifying revenue away from the mature US core. The CP+ data advantage widens as flow grows, automation handles an ever-larger share of trades at high margin, and the indexation wave sends systematic and ETF flow onto the platform's closing auctions. In this world the soft Q3 2025 looks like a cyclical blip in a re-accelerating growth story, operating margin expands as the new protocols scale on a built-out platform, and the accelerated buyback steadily compounds per-share value. The company stops being seen as an ex-growth incumbent and re-rates as a multi-protocol fixed-income infrastructure franchise.

Base case. Management delivers roughly what it has guided. US credit, the mature core, grows in line with overall electronification (low-to-mid single digits, with the occasional soft quarter when volatility and spreads compress), while the newer initiatives - block, portfolio trading, Mid-X, automation, and international - continue to generate around half of incremental revenue and keep total growth in the low double digits in good years and mid-single digits in quiet ones. Tradeweb remains a persistent share-taker in portfolio trading but does not break MarketAxess's hold on Open Trading and US credit RFQ. Expenses grow at the guided ~8% as the company invests through the build-out, keeping margins healthy but not dramatically expanding. The dividend keeps rising ~3% a year and the buyback keeps shrinking the share count. The result is a steady, cash-generative compounder whose growth rate is respectable but no longer explosive, with results that swing quarter-to-quarter with credit-market volatility.

Bear case. The maturing US core bites harder than expected just as Tradeweb reaches an inflection in credit. Multi-homing clients route more and more credit flow to Tradeweb's improving liquidity, and Trumid simultaneously peels off the high-value anonymous flow that Open Trading depends on. Because the model is volume-and-share driven, even modest share losses compound: less flow means worse pricing data, which means a weaker CP+ edge, which means still less flow. Meanwhile the new protocols (block, Mid-X, new-issue) ramp slower than the marketing implied, so they fail to offset the eroding core, and the company is left spending ~8% more on expenses each year against decelerating revenue, compressing operating margin. A prolonged low-volatility credit regime, like the one that produced the 1% revenue quarter in Q3 2025, persists rather than passing, turning a single soft quarter into a soft year. The stock de-rates from a growth franchise to a mature, share-losing exchange whose buyback is the main thing supporting per-share results.

A note on Section 13: I searched SemiAnalysis, Stratechery, and MBI Deep Dives for qualifying MarketAxess coverage. SemiAnalysis and Stratechery focus on semiconductors and consumer/platform tech and have no deep dive on MarketAxess; MBI Deep Dives has not published a deep dive primarily about MKTX within the relevant window. Per the empty-case rule, the Further Reading section is omitted entirely rather than padded with a placeholder.

Sources: MarketAxess Q1 2026 8-K (SEC) · Q1 2026 transcript (Motley Fool) · Q1 2026 results (Businesswire) · Q4 2025 transcript (Benzinga) · Q3 2025 transcript (Insider Monkey) · Q2 2025 transcript (Insider Monkey) · Q1 2025 transcript (Seeking Alpha) · MarketAxess history (Wikipedia) · Richard McVey (Fortune) · Coalition Greenwich US corp bond e-trading 2024 · Tradeweb (Wikipedia) · Tradeweb market cap (Macrotrends) · ICE market cap (companiesmarketcap) · CME market cap (companiesmarketcap) · MKTX dividend/buyback 8-Ks (SEC)

A few honest caveats on what I delivered: I produced the report inline (this environment exposes only web tools, so I could not write a .md file directly - copy this into a .md to save it). Revenue-type mix percentages are approximate and rounded from disclosure patterns rather than a single cited line item, and per the report rules I deliberately avoided absolute dollar revenue/margin figures in the narrative. The insider section rests on the MoatMap spine plus a recency cross-check; all four transactions are immaterial. Want me to expand any section (e.g., a deeper competitive teardown of Tradeweb, or a fuller protocol-by-protocol product map)?