In September, total digital-asset market capitalization recovered to about $2.84 trillion, up roughly 8% month over month; average daily spot volume rose about 33% MoM, and DeFi-sector market cap about 25% MoM. While price and turnover warmed, the combined market value of crypto-related equities tracked by publicly cited research fell about 2%, with Circle (CRCL) down about 14% that month. A repair in spot volume does not automatically repair listed-company valuations. Markets trade “on-chain activity”; capital scores “what can land in the P&L.”
A deeper mismatch: volume turns into trading fees more readily; payments, custody and network services still pass through take rates, revenue share and operating cost before they become profit. DeFi total value locked (TVL) returned to about $94.7 billion, still about 42% lower year over year. For listed companies, the real question is not only whether the chain is busy, but how money enters their own statements after the buzz — at what pace, at what gross margin, and how dependent on market beta.
Three Chains, Three Entry Points
Coinbase, Robinhood and Circle are each building a chain, but from different starting points. One leans toward a trading and lending ecosystem, one toward migrating retail securities users, one toward institutional stablecoin settlement — different entry points, different monetization clocks.
Coinbase’s Base is more mature, with trading, lending and stablecoins already coexisting. Robinhood’s Robinhood Chain advantage is securities-account users and distribution. Circle’s Arc centers on USDC institutional payments and settlement. Base’s TVL still leads by a wide margin — reportedly about six times Robinhood Chain’s — while the latter has been more active recently on active addresses. Arc is newest, with broader institutional participation; payment volume and fee monetization remain early.
Coinbase: Turning Base Ecosystem Depth Into a Funnel
Base launched in August 2023, with daily transactions around 9 million and about $4 billion of USDC on-chain; TVL on Morpho already exceeds $4 billion. In July the company further stressed Base as financial infrastructure covering lending, payments and tokenization. According to publicly cited research estimates, Base’s quarterly direct net revenue sits roughly in a $5–25 million range, mainly from sequencer fees — annualized about $20 million to $100 million — still small versus Coinbase’s overall business, yet able to feed the exchange, USDC and institutional funnel. In other words, Base today looks more like an entry and scene layer than a profit center that can carry valuation alone.
In the second quarter of 2026, Coinbase’s digital-asset trading share was about 10.3%; subscription and services revenue about $555 million, roughly 48% of net revenue; average USDC balances on its products about $20 billion. AI-agent payments remain early: reportedly more than 97% of on-chain agent transaction counts used x402, and more than 90% of agent stablecoin volume occurred on Base.
Over the same period, total revenue was about $1.22 billion, down about 19% year over year; GAAP net loss about $360 million; adjusted EBITDA about $208 million; adjusted net loss excluding fair-value moves on digital assets and investments about $105 million. The narrative emphasis is ecosystem depth: customers can enter USDC, Base and institutional rails — long-term value larger than the tens of millions to $100 million sequencer scale. But a second-quarter adjusted loss still shows growth has not fully converted into operating profit.
Robinhood: A Securities-User Funnel Harder Than Any Single Chain
Robinhood Chain launched in July; in about three months TVL surpassed $1 billion. Early daily transaction peaks near 20 million later settled around 6 million; DEX daily volume moved from more than $4 billion toward a roughly $10 billion scale at times. Some of that flow was meme-driven and volatile — early peaks should not be annualized straight.
HOOD’s harder edge is the securities-account funnel: stocks, options, prediction markets and crypto can land on the same user base. Users need not become “on-chain natives” first to be brought into tokenized and on-chain products. According to CoinDesk, September volume on Robinhood’s tokenized-stock trading venue was about $6.57 billion, up about 407% MoM, with roughly 42% share. That is not the same metric as on-chain turnover, but it shows tokenization execution — distribution is often harder to copy than a single day’s TVL.
Second-quarter total net revenue was about $1.308 billion, up about 32% YoY; net income about $573 million; adjusted EBITDA about $741 million; funded customers about 28.4 million; platform assets about $369 billion. Crypto trading revenue fell about 38% YoY, yet overall still grew 32%, with options and securities offsetting. Reportedly about 13 business lines have reached more than $100 million in annualized revenue. Sequencer share: retain 50% initially, then 70% after a cumulative threshold, then 85%. According to publicly cited research, third-quarter net revenue from that source may be around $25 million.
Among the three, HOOD’s product and execution look the most complete; a multi-product mix also cushions profit when crypto is soft. But a price-to-sales (PS) multiple around 20× already prices growth, and early speculative volume peaks cannot be treated as the steady state.
Circle: Beyond USDC Scale, Moving Past Interest Dependence
Circle’s second-quarter total revenue and reserve income combined were about $701 million, of which reserve income about $668 million and other about $34 million. USDC in circulation rose, but reserve yield fell to about 3.5%, down roughly 66 basis points: average USDC up about 25%, reserve income up only about 5%. Distribution costs run high: second-quarter distribution and transaction-related costs about $412 million, leaving about $289 million, a retention of about 41%; adjusted EBITDA about $143 million.
Arc’s mainnet went live on September 16 as a Layer 1, using USDC for gas and plugging into CCTP; participants include BlackRock, Visa, Mastercard, DTCC and others. Weeks after launch, TVL exceeded $500 million, of which about $490 million sat on Morpho and Aave — still mostly lending; payment monetization comes later. Genesis issued 100 billion ARC tokens, with a path toward proof of stake; arrangements for publicly traded ARC remain undecided. The narrative is an institutional payment and agent-trading stack of USDC, CCTP and Arc: valuing only on reserve interest understates the payment-network option; valuing as a high-growth payments platform still lacks proof of stable non-interest profit. For investors, the key watchpoint is not “whether big institutions show up,” but whether payment and settlement fees can gradually stand apart from interest income.
Valuation Separates Profit Engines, Not TVL Rankings
Rough PS multiples: HOOD about 19.5×, COIN about 7.5×, CRCL about 7.1×. Behind the numbers sit different profit engines. Setting share prices aside for a moment, business model, product breadth and user funnel lean toward HOOD; COIN at a lower PS has room for mean reversion and profit repair, but needs operating profit to show up; CRCL is better studied by splitting USDC scale, channel share and payment fees rather than interest spread alone. The point of a valuation compare is not to find “who is cheapest,” but to see how much premium the market pays respectively for a trading ecosystem, retail distribution and a stablecoin network.
On competition, HOOD must prove growth can carry the multiple; COIN must turn ecosystem depth into profit; CRCL must deliver non-interest growth. Ranking only by on-chain volume or TVL flattens three different paths onto one board. The next stage of on-chain finance may no longer be who launches first, but who can steadily convert traffic, users and institutional relationships into repeatable fee rights.
Put the three together, and a more useful question may be: who is selling trading, who is selling distribution, who is selling a dollar network. Once that answer is clear, lining up PS, TVL and quarterly profit is less likely to be pulled off course by a single metric.
This article is for informational purposes only and does not constitute investment advice. Market data and research estimates change over time; please verify the latest public disclosures before citing.
