Why Crypto Firms Grow but Struggle to Profit

In brief: Coinbase, Circle, and Robinhood have each found fresh ways to grow, in stablecoins, payments, and prediction markets, but their latest quarterly results show those lines do not yet shield earnings from the swings of a cyclical trading market. Coinbase's third-quarter 2026 revenue fell about 35 percent to $1.22 billion even as subscription and services reached 49 percent of net revenue, and Circle's reserve income dropped 8 percent as a one-off token presale flattered the quarter. The lesson for institutions is that diversification has changed the revenue mix without yet changing the cycle.

The companies at the center of the digital-asset industry have spent two years answering a single criticism: that their fortunes rise and fall with crypto prices. They diversified. Coinbase and Robinhood pushed into adjacent financial services, while Circle bet that dollar-backed digital money would become a routine way to move value. Those efforts gave each firm more ways to earn. According to CoinDesk's November 2026 analysis, they have not made the business of crypto any less cyclical.

The tension is simple to state. New revenue streams are real, and some are large, but they rest on foundations that move with interest rates, trading volume, and sentiment. A company can diversify its product line and still find that every line leans on the same weather.

What did Coinbase's latest quarter actually show?

Coinbase reported a $719 million net profit for the third quarter of 2026, or $2.51 per diluted share, yet that headline flattered the underlying business. As Yahoo Finance reported, revenue fell about 35 percent from a year earlier to $1.22 billion as consumer spot trading dropped sharply. The profit beat came largely from gains on the company's own crypto holdings rather than from running the exchange.

The margin picture makes the point in plain numbers. Trefis calculated that Coinbase's operating margin fell from 23.8 percent a year earlier to 2.5 percent in fiscal Q3 2026, and most of the quarter's GAAP earnings per share came from unrealized investment gains against a consensus that had projected a loss. Strip out the mark-to-market on the balance sheet and the operating engine was close to flat.

Diversification did move the mix. According to Investing.com, subscription and services reached a milestone 49 percent of net revenue in the quarter, roughly half the business. That category includes stablecoin income, staking, and custody, and it has grown steadily: Coinbase's stablecoin revenue alone rose 48 percent to about $1.35 billion in 2025, disclosed in its annual report. The problem is that the stablecoin line is itself a function of interest rates on reserve assets, which is to say it trades one kind of cyclicality for another.

Why did Circle's results flatter a weaker core?

Circle, the issuer of USDC, posted third-quarter 2026 revenue and reserve income of about $865 million, a number that looked healthy until you read the composition. As The Block reported, the quarter also included a 5 percent workforce reduction, an unusual move for a company supposedly riding a growth wave.

The detail that matters is what lifted the top line. Circle recorded $182 million from the presale of its ARC token, a one-time event, while its core reserve income fell 8 percent, according to FXStreet. In other words, a one-off sale masked a decline in the recurring business that actually defines Circle. That recurring business is the yield Circle earns on the Treasuries and cash backing USDC, and it compresses when the Federal Reserve cuts rates. Circle's economics are, at root, a bet on where short-term rates sit.

This is the structural catch for stablecoin issuers generally. The model is elegant when rates are high and punishing when they fall. Tether, the largest issuer, cleared more than $10 billion in profit through the first nine months of 2025 on much the same mechanism, so the upside is genuine. But a business whose central driver is the Fed funds rate is not diversified away from macro risk. It is exposed to a different macro risk than trading volume, which is not the same thing as being insulated.

How is Robinhood faring with prediction markets?

Robinhood offers the clearest example of a new line doing exactly what it was meant to do, and of the ceiling on how much that helps. In its third-quarter 2026 report, the company's crypto revenue slumped while its prediction-market, or event-contract, revenue climbed sharply. As TheStreet put it, prediction-market revenue soared as crypto slumped, which is diversification working as advertised.

Yet the market read the quarter unkindly. CNBC noted that Robinhood shares fell as the crypto slump weighed on profits, offsetting gains in the newer businesses. The event-contract line grew fast enough to cushion the fall but not fast enough to reverse it. One cyclical business, crypto trading, was partly offset by another fast-growing but still-young line whose own durability through a full cycle is untested.

Is diversification actually reducing the cycle?

Here the pattern across all three firms comes into focus, and it is less reassuring than the growth headlines suggest. Each company added revenue that is real and, in some cases, substantial. None has yet added revenue that is genuinely counter-cyclical to the whole crypto complex. Stablecoin reserve income tracks interest rates. Trading tracks volume and sentiment. Event contracts track their own, still-forming demand curve. When markets sold off in late 2026, with Bitcoin steadying after an AI-led selloff and more than $1 billion in bullish bets liquidated, every one of these lines felt it at once, directly or through the knock-on effect on customer activity.

That correlation is the quiet story in the earnings. A diversified revenue mix is only defensive if its parts respond to different shocks. The crypto majors have assembled portfolios of businesses that, so far, still respond to the same ones. The mix has changed; the beta has not.

There is a wider shakeout running underneath all of this. CoinDesk reported in August 2026 that more than 100 crypto projects folded during the year, a dot-com-style culling that separates firms with durable economics from those without. The survivors are precisely the companies now diversifying hardest, which is why the question of whether that diversification works is not academic. It is the test of which business models outlast the cycle.

What to watch next

The near-term signal is interest rates. If the Fed continues easing, stablecoin reserve income compresses across the board, and the subscription-and-services lines that now carry roughly half of Coinbase's revenue will show the strain. Watch whether issuers like Circle can replace reserve yield with fee-based revenue from payments and settlement, the use case that would make a stablecoin business genuinely less rate-dependent. ARC presale money will not recur; fee income from actual transaction volume would.

The second signal is whether the newer lines keep growing when crypto is flat rather than falling. Robinhood's event contracts have been tested in a down market but not in a long, dull one, where neither fear nor greed drives activity. A revenue stream that holds up in a quiet market is worth far more than one that merely spikes in a volatile one.

The open question, then, is not whether the crypto industry can grow. It plainly can, and in more directions than before. The question is whether any of these firms can build a line of business that earns steadily when prices are not moving at all. Until one of them does, the sector remains what the numbers say it is: better diversified than it was, and still waiting for the cycle to decide its earnings. For institutions weighing exposure to the programmable, auditable instruments these firms are building, that distinction, between more revenue streams and more resilient ones, is the one worth underwriting.