Circle Earnings Face Rate Risk as KelpDAO Exploit Shakes DeFi

Circle earnings are getting a fresh reality check just as DeFi absorbs one of its ugliest liquidity shocks of the year. A new Coin Metrics report published this week lays out, in uncomfortable detail, how the stablecoin issuer's $2.7 billion revenue engine leans almost entirely on interest rates it does not control, while a $290 million exploit on KelpDAO shows how fast a single misconfigured bridge can drain billions from lending markets. Two stories, one theme: the plumbing under crypto is more fragile than the quarterly numbers suggest.

Circle's $2.7 Billion Revenue Engine, Explained

Here is the short version. Circle, trading on the NYSE under the ticker CRCL, wrapped fiscal year 2025 with about $2.7 billion in revenue, a 64% jump from the prior year. Almost all of it came from sitting on USDC reserves and collecting yield. USDC's circulating supply hit $75.3 billion by year-end, holding a steady 28% slice of the stablecoin market. Adjusted transfer volumes jumped roughly 320%, which sounds explosive until you realize the company barely monetizes that activity yet.

Subscription and service fees chipped in $110 million. That is not nothing, but it is small. Reserve income still accounted for 96% of the top line. After Circle paid its distribution partners their cut, revenue less distribution costs, the figure Wall Street watches, landed at $1.08 billion, a 39% margin. Detailed quarterly breakdowns are available through Circle's Circle earnings investor portal, which shows the shape of that reliance in granular form.

Good year. Dangerous structure.

What Happens If the Fed Cuts Rates?

This is where Circle's story gets interesting, and where bulls and bears part ways. The Coin Metrics analysts used FY2025 as their baseline, roughly a 4.1% effective yield on an average $65 billion supply, and then ran the scenarios that actually matter to a shareholder. A 100-basis-point drop in yields would chop 25% to 30% off annual revenue. That is a brutal sensitivity for a company whose stock trades on growth multiples.

The modeling gets uglier as you move down the curve. If sustained rates stay at 4% to 4.5% and USDC supply climbs to $90 billion to $110 billion, reserve income could push past $4 billion. Moderate easing to 3% to 3.5%, which lines up with current Fed projections, would mostly wash out if supply keeps expanding. But a deeper cut to around 2.5% could slash income by 30% to 50% even in a strong growth scenario. The math is blunt: Circle is essentially a used bet on the Fed not being dovish.

• Base case FY2025: ~4.1% yield on $65B average supply
• Bull case: 4% to 4.5% rates, supply to $110B, income above $4B
• Neutral: 3% to 3.5% rates, supply growth offsets yield compression
• Bear case: 2.5% rates, income down 30% to 50% even with growth

The Coinbase Deal Everyone Should Be Watching

Buried inside the report is the single most important line item nobody talks about at cocktail parties: the August 2026 renewal of Circle's revenue-sharing arrangement with Coinbase. Under current terms, Coinbase gets all the interest on USDC held on its platform, plus half the interest earned everywhere else. In 2025 that arrangement captured roughly 51% of gross reserve income. Half the yield pie walks out the door before Circle books a dollar.

A better deal in August, with more diversified distribution partners, would widen Circle's take-rate and cushion the blow if rates fall. A weaker deal, or a status quo deal in a falling-rate world, would stack two hits on the same quarter. The report also flags the CLARITY Act, a piece of pending US legislation that could indirectly restrict the kind of yield-sharing programs stablecoin issuers have built their economics around. New revenue lines like the Circle Payments Network and the Cross-Chain Transfer Protocol, CCTP, are still tiny but at least point toward a future where Circle is not just a money market fund with a logo.

Circle's future is a balancing act between rate-sensitive reserve income and growing usage-driven revenue.

KelpDAO Exploit: $290 Million Gone, Then the Bank Run

Now for the ugly half of the report. On April 18, an attacker exploited a misconfigured LayerZero decentralized verifier network, a DVN, inside KelpDAO's rsETH cross-chain bridge. The result was roughly 116,500 unbacked rsETH minted out of thin air, about 18% of total supply. LayerZero's own incident statement covering the KelpDAO exploit details how the misconfiguration slipped past verification and why it mattered across chains.

The attacker did not just dump the fake tokens. They used them as collateral, mostly on Aave, to borrow WETH. That single move pushed Aave's v3 WETH market to 100% utilization almost instantly. Lenders panicked. Utilization rates spiked across USDC and USDT pools, interest rates on those pools went vertical, and more than $9 billion in deposits were yanked from DeFi lending protocols in what Coin Metrics correctly calls a bank run. Classic cascade.

This is the part that should keep restaking believers up at night. The exploit was not a flaw in Aave. It was not even really a flaw in KelpDAO's core protocol. It was a configuration error on a bridge component, and it still managed to drain nine figures and scare ten figures out of an entirely separate lending market. That is not a bug. That is the design.

Why Does This Matter for Crypto Investors?

Because both stories say the same thing in different languages. Circle's earnings are strong today and structurally exposed tomorrow. DeFi's yields are attractive today and structurally fragile tomorrow. If you are holding CRCL, you are short the Fed and long the Coinbase renegotiation. If you are farming rsETH or parking stables in Aave, you are one misconfigured DVN away from a bank run you did not see coming.

The full Coin Metrics report walks through the modeling and the on-chain forensics in more depth, including the collateral curation failures that let unbacked tokens flow into a blue-chip lending market without friction. The research team's own framing is careful, but the implication is not subtle: layered DeFi, where bridges, restaking derivatives, and lending protocols all lean on each other, is exactly as strong as its weakest misconfigured component.

Investors have spent 2026 celebrating stablecoin adoption and restaking yields as if the two trends were unrelated. They are not. USDC sitting in a Coinbase wallet and rsETH borrowed against on Aave are parts of the same on-chain economy, and when one piece cracks, the other feels it within minutes. Circle's $75.3 billion supply is a feature of that economy. So was KelpDAO's bridge, until it was not.

What to Watch Next

Three things. First, the Fed's next meeting and any signal on the path toward 3% or lower. Second, anything that leaks about the Circle and Coinbase renewal, which is now less than four months away. Third, whether other restaking protocols and bridges publish audits of their DVN configurations before someone repeats the April 18 playbook on a bigger target. The Coin Metrics team stops short of a prediction. We will not.

If rates fall faster than USDC grows, and if DeFi keeps patching its plumbing one exploit at a time, the next twelve months will be less about who is building the future of finance and more about who is still standing when the next configuration error hits production.