Coinbase’s second quarter showed a familiar problem: when retail traders back off, the exchange feels it fast. Revenue fell, trading volumes slumped, and the company leaned harder on stablecoins and subscription products to keep the lights bright.
- Third straight quarterly GAAP loss
- Trading revenue and spot volume fell again
- USDC and subscriptions are carrying more weight
- Coinbase is still hunting for a post-trading growth engine
[Coinbase posts third straight quarterly loss as retail] reported a third consecutive quarterly loss, with Q2 revenue at $1.22 billion for the three months ended June 30 and net loss just under $360 million. That followed a $394 million net loss in Q1, despite Coinbase previously posting a $1.43 billion profit in Q125.
The headline is simple: fewer traders, fewer trades, fewer fees. Coinbase still makes its core money from transaction activity, and when the market gets sleepy, the results get ugly.
Monthly transacting users fell to 7.6 million from 8.2 million in Q1 and 8.7 million in Q225. Transaction revenue dropped 21% to $599.2 million. Consumer transaction revenue came in at $451.7 million, down from $567 million in Q1, while institutional transaction revenue fell 26.3% to $100 million. “Other” transaction revenue added $47.4 million.
Spot trading volume also slid, landing at $146.4 billion, down 24% quarter over quarter. Consumer spot volume was $25.8 billion, down 23.7%, and institutional spot volume was $120.6 billion, also down 24%.
Coinbase did not include its usual token-by-token breakdown of transaction volume and revenue in the Q2 report. That matters because it normally helps analysts see which assets are driving activity and whether Coinbase is quietly becoming more dependent on a handful of tokens. Less detail is rarely a sign that management is thrilled with the picture.
Still, Coinbase said its crypto trading volume market share hit an all-time high of 10.3%, up 1.2 points from Q1. That sounds good until you remember the obvious caveat: a bigger slice of a shrinking market is still a shrinking market. Market share is useful. Market share in a dead market is just a nicer-looking obituary.
[Understanding the Basics of HTML Content Processing] Glassnode analysts said BTC spot trading volume “has fallen to its lowest since 2019, ” which helps explain why Coinbase’s trading numbers look weak. This is not just a Coinbase problem. Bitcoin and broader crypto trading have cooled, and the retail crowd that once chased every green candle seems to have taken a long coffee break.
Robinhood told a similar story. The company reported about $100 million in crypto trading revenue for Q2, down 38% year-on-year, while its prediction market brought in $156 million over the same period. That is not an apples-to-apples comparison, but it does reinforce the same point: consumer crypto trading is no longer the easy growth lever it once was.
So Coinbase is doing what mature exchanges eventually do when trading gets choppy: it is trying to look less like a casino and more like a platform. Its Subscription & Services segment brought in $555.2 million in Q2, down nearly 5% from Q1, and represented 48% of overall revenue, up seven points from Q1.
For readers new to the term, Subscription & Services is Coinbase’s non-trading revenue bucket. It includes stablecoin revenue, blockchain rewards, and other product-related income. In plain English: it is the part of the business that does not depend on traders waking up in a mood.
Within that segment, stablecoin revenue was $292.1 million, down about 5%, and blockchain rewards were $83.3 million, down more than 42%. Coinbase also said [USDC (cryptocurrency)] holdings in its products hit a new all-time high.
That USDC exposure is the real story under the hood. Coinbase CEO Brian Armstrong said the company is a “multi-stablecoin platform” and wants to “provide the stablecoins that all of our customers want to use.” He added that Coinbase wants to “strike economic deals with every major stablecoin out there.”
That is pragmatic, and it is also a bit of a confession. Coinbase is increasingly dependent on the stablecoin plumbing layer of crypto, not just speculative trading. USDC, the dollar-pegged stablecoin issued by Circle, has become central to Coinbase’s economics.
Alesia Haas said Coinbase has “already met the conditions for the Circle contract to renew. So it will renew on the same terms … We will continue to work on growing USDC, partnering with Circle, and driving that ecosystem.” She also said the point was to “take away any ambiguity.” The current Coinbase-Circle USDC deal expires in August.
That renewal matters because stablecoin revenue is no side hustle for Coinbase anymore. If the terms stay the same, Coinbase preserves a valuable revenue stream. If they change in the future, the company’s economics could shift again. Stablecoins are one of crypto’s clearest real use cases, but concentration risk does not disappear just because the asset involved has a neat ticker and a clean dollar peg.
Coinbase is also trying to build outside the old transaction-fee model. The company said its prediction market product generated “$100M+” in annualized revenue in Q2 and was up 106% from Q1. Annualized revenue means a run-rate estimate based on current activity, not actual quarterly revenue, so readers should not confuse marketing math with hard cash already in the bank.
Coinbase One also passed one million members, giving the company another recurring-revenue angle to push. On paper, that is exactly what a slower trading business wants: more subscriptions, more predictability, less dependence on traders doing trader things.
Expenses totaled $1.33 billion in Q2, down 7% from Q1. Coinbase also reported $52.4 million in restructuring costs tied to laying off 14% of its workforce earlier this year. Stock-based compensation came to $238 million in Q2, and the company expects $245 million in Q3.
Those numbers matter because Coinbase cannot out-muscle a weak market with wishful thinking. If trading activity stays soft, the company needs lower costs and stronger recurring revenue to avoid becoming a very expensive way to learn that retail speculation has a pulse, but not always a strong one.
There was also the usual round of leadership churn. Chief legal officer Paul Grewal is leaving after six years and will transition to an advisory role at the end of the month while continuing to serve on the board of Coinbase National Trust Company. Reuters reported that Molly Abraham will become general counsel, and Ryan VanGrack was elevated to vice-chair and head of corporate affairs.
Senior engineer Brock Miller also left after eight years to join Anthropic, and on July 28 Coinbase named Rob Witoff as its new CTO. None of that is unusual in a company Coinbase’s size, but the steady motion says something about the business: it is still adapting, still reorganizing, still trying to find the shape of its next phase.
Operationally, Coinbase still has work to do. Its most recent unplanned downtime came on July 14 and lasted nearly an hour. In June, Base, Coinbase’s Ethereum layer-2 network, suffered two block production issues in as many days.
For newcomers, Base is a blockchain network built on top of Ethereum that is meant to make transactions cheaper and faster. A layer-2 like Base can be useful, but only if it works reliably. If you want users to trust you with trading, custody, and onchain activity, downtime is not a quirky footnote. It is a problem.
Base has also been the site of some awkward product experimentation. Jesse Pollak said he had handed the Base app back to the Coinbase “mothership” and that the belief crypto adoption would come from creators, content, and messaging had been “the wrong bet.” He also said, “just put your head down and build. so that’s what i’m doing.”
Armstrong was even more blunt, saying “content coins” “didn’t work and we pivoted early this year. We messed up, time to turn the page.” That is refreshingly honest. A lot of crypto operators would rather keep selling the same failed storyline until the last sucker leaves the room.
The broader lesson is simple: tokens tied to creators and content mostly did not become a serious adoption wedge. They became another speculative gimmick with better branding. Coinbase and Base deserve credit for pivoting, but the episode also shows how easy it is in crypto to confuse attention with product-market fit.
Coinbase is now pushing a different onchain angle. On July 27 it announced a DEX “Launches” feature for Base or Solana tokens as soon as they are live onchain. A decentralized exchange, or DEX, lets users trade directly onchain using smart contracts rather than depositing assets with a centralized company. That can be a real step toward more open markets, or just a faster lane for token churn if the incentives are sloppy.
The company is also leaning into AI as a way to ship faster. Coinbase said “roughly 40%” of its code was being written by or with AI assistance this spring, and this month it said “roughly 100%” of “everything merged into our codebase” is AI-generated but human-reviewed.
That sounds impressive, but it needs context. AI-assisted coding can absolutely improve productivity when humans are still reviewing what gets shipped. It can also become a management stat that looks shinier than the actual product if reliability issues keep cropping up. Given the recent downtime and Base hiccups, Coinbase still has to prove that the AI story is helping execution, not just decorating a slide deck.
Coinbase shares closed Thursday at $163.58, up 2.2%, though the stock was still down nearly 28% since the start of the year and more than 56% over the past 12 months. Shares fell as much as 7% in after-hours trading after the report.
The market is telling Coinbase something fairly direct: the old trading boom is not guaranteed, and a “platform” story is only as good as the actual revenue it produces. Stablecoins, subscriptions, prediction markets, DEX tools, and AI-assisted engineering may help. They may even help a lot. But none of them erase the core truth that Coinbase still lives and dies by crypto activity.
That is the good news and the bad news. Coinbase is trying to build a sturdier business. The bad news is that the foundation is still tied to a market that gets quiet fast when the crowd stops cheering.
Key questions and takeaways
-
Why did Coinbase post another loss?
Trading activity weakened across the crypto market, especially among retail users. Lower volume means lower transaction fees, and transaction revenue is still Coinbase’s biggest engine. -
Why does USDC matter so much?
USDC is now central to Coinbase’s stablecoin revenue and broader product mix. Even with the Circle deal set to renew on the same terms, Coinbase’s dependence on USDC is a real concentration risk. -
Is a higher market share a big win?
Only if the overall market is healthy. Coinbase’s 10.3% crypto trading volume market share looks strong, but it matters less if total trading activity keeps shrinking. -
Can subscriptions replace trading revenue?
They help diversify income, and Coinbase One passing one million members is encouraging. But subscriptions are unlikely to fully replace trading in a weak market without much stronger growth. -
What went wrong with creator coins?
Coinbase and Base admitted the thesis was flawed. The bet that creators, content, and messaging would drive adoption did not land, which is a useful reminder that hype is not the same thing as usage. -
What does the new DEX feature change?
It gives users faster access to new Base or Solana tokens directly onchain. That could be useful for discovery, but it could also add another speculative churn machine if poorly managed.
Further reading
A few useful documents and related reports for anyone tracking Coinbase, USDC, and the shifting exchange revenue mix.
- Understanding HTML Content Processing for Title Extraction
- Error extracting content
- EX-10.1
- Coinbase posts surprise loss on crypto trading slowdown
- Coinbase Launches USDC Vault With Ethena and Morpho
- Circle Moves 4.4B USDC to Coinbase in Record HyperEVM
- Coinbase, MassPay Partner for USDC Cross-Border Payouts
