Volume without intent is just digital noise.
Everyone thinks stablecoins are the future of payments. Coinbase's Head of Stablecoins, Brian Foster, recently claimed that stablecoin transaction volume will surpass traditional fiat within five years. The data says a different story.
Hook
In Q2 2025, on-chain stablecoin transfer volume hit $4.2 trillion on Ethereum alone. Yet, when I filtered for transactions under $100—the sweet spot for everyday payments—the number dropped to $12 billion. That's 0.28% of the total. The rest? Arbitrage, exchange internal sweeps, wash trading. Volume without intent is just digital noise.
Context
Foster's prediction is not new. It's a rehash of the "crypto payment" narrative that has been recycled since 2017. But now, with institutional backing and regulatory tailwinds, the market is buying it. The bull market amplifies euphoria. Projects like Solana and Base are positioning themselves as payment rails. But the assumption that stablecoins will replace Visa and Mastercard in five years ignores a fundamental truth: the data doesn't support it.
Based on my experience auditing ICO contracts in 2017, I learned to spot when marketing claims hide technical flaws. This prediction is no different. Let's go on-chain.
Core: The On-Chain Evidence Chain
First, let's look at active addresses. Stablecoin users on Ethereum peaked at 1.2 million unique addresses per day in March 2025. Visa processes 3.5 billion transactions per day. Even if every single active address made one payment—which they don't—it's still orders of magnitude below.
Second, median transfer size. USDC's median transfer size on Ethereum is $1,350. On Solana, it's $420. That's not coffee money. That's institutional settlements and DeFi margin calls. Real-world payment volumes are dominated by sub-$50 transactions. Until stablecoins are used for those, the comparison is fraudulent.
Third, velocity. I built a dashboard in 2020 to track yield farming mechanics—today I use it to track stablecoin circulation. The average USDC sits in one wallet for 2.3 days before moving. That's a trading instrument, not a medium of exchange. Compare to cash: the average dollar bill changes hands 4 times per day. Stablecoins are slow money.
Fourth, the distribution. Top 100 wallets hold 68% of USDC supply. That's not a payment network. That's a concentration of capital. Payments require dispersion—millions of addresses with small balances. We don't have that.
Contrarian: Correlation Is Not Causation
The bull case argues that stablecoin volume is growing exponentially. It is. But growth in trading volume is not growth in payment volume. The two are conflated. If I transfer $10 million USDC from Binance to Kraken to execute an arbitrage, that's counted as "volume." It's not a payment. It's noise.
In 2021, I exposed wash-trading on NFT marketplaces—clustered wallets generating fake volume. The same logic applies here. The majority of stablecoin transactions are non-economic: spam airdrops, dusting attacks, flash loans. Until we strip out these artifacts, the raw volume number is meaningless.
Moreover, the infrastructure for consumer payments is still broken. Gas fees on Ethereum L1 make a $3 coffee cost $0.50 in fees. L2s help, but then you need bridges, which add latency and security risks. The user experience is still terrible. Why would a merchant accept crypto when it takes 20 seconds to confirm a transaction? Visa does it in milliseconds.
Takeaway: Watch the Signal, Not the Noise
The next six months will tell us whether stablecoins are really moving toward payments. I'm watching three metrics: (1) number of daily on-chain transactions under $50, (2) share of volume from retail wallets (balance < $1,000), (3) merchant adoption—actual point-of-sale integrations, not just exchange listings. If these don't break out by 2026, the five-year prediction is dead.
Until then, treat Coinbase's forecast as marketing, not analysis. Follow the gas, not the gossip. Volatility is the tax on ignorance. And in this case, ignorance is believing that exchange volume equals payment volume.
