Stablecoins drain $10B in eight weeks while crypto's liquidity crutch goes quiet

More than $10 billion has left the aggregate stablecoin complex since May, according to a market update circulated by Cointelegraph on 12 July 2026. The total market capitalisation of dollar-pegged tokens, which had hovered near record highs through early spring, has given back a slice large enough to reset the assumptions under which the last crypto bull market was financed. The outflow is the first clean, dated signal that the bid beneath the cycle is changing shape, not just stuttering.

The story matters because stablecoins have, for the past two years, functioned as the marginal liquidity layer for crypto. Issuance into the system is, in effect, a synthetic dollar-deposit expansion aimed at exchanges, market-makers and offshore desks; redemption is the reverse. When the float contracts, the easiest explanation is that risk is being moved off the table, and the harder one, more structurally important, is that the offshore dollar plumbing is being rerouted into something else.

The number, properly read

Cointelegraph's update framed the move with two anchors: an absolute figure and a comparison set. The figure is the $10 billion-plus contraction in stablecoin market capitalisation since May. The comparison set is the broader risk tape: bitcoin's monthly relative-strength index has drifted to a level last seen in 2022, the same date stamp Cointelegraph flagged on 11 July 2026.

Read together, the two reads say something specific. In 2022 the contractive phase was driven by a forced unwind of leveraged long positions, the collapse of a large algorithmic stablecoin and a Federal Reserve rate-hike cycle that pulled dollar funding back to the United States. This episode so far looks different in shape. There is no algorithmic-coin blow-up in the public record. The Federal Reserve is, depending on the meeting you pick, either cutting or on hold, not tightening. The stablecoin float is shrinking into a tape that is generally richer in collateral and less levered. That is consistent with capital being rotated out of the crypto complex into adjacent risk: US equities, private credit, the AI-infrastructure capex surge that has pulled forward nearly 187% more corporate spending on physical AI buildout in the past twelve months than the comparable twelve-month window before.

That rotation does not require anyone to be bearish on crypto. It only requires US equities, AI-adjacent infrastructure, or simply short-dated US Treasuries to become a better risk-adjusted home for marginal dollar. The stablecoin contraction is the receiving end of that choice.

The capex gravity well

The AI-infrastructure line in the Cointelegraph update deserves more weight than it usually gets in crypto coverage. A 187% year-on-year surge in physical-AI-infrastructure capex is not a financial-asset event; it is an industrial event. It absorbs data-centre construction, power-purchase agreements, grid interconnects, custom silicon, and the heavy electrical equipment that the Western press has spent two years writing about. In a world of finite institutional risk budgets, every dollar committed to a multi-year hyperscale build-out is a dollar not committed to a Bitcoin-denominated treasury strategy, a stablecoin-minting trade, or an altcoin venture round.

This is the part of the story the bear case for crypto tends to mis-describe. The outflows are not, on the evidence so far, the consequence of a credit shock, a regulatory shock, or a counterparty failure. They are the consequence of competing bid: a corporate capex cycle that the AI trade has made artificially large, draining the marginal dollar that would otherwise have washed through Tether, Circle and their smaller rivals into exchange liquidity. Stablecoins are still growing on a year-over-year basis; they are merely growing more slowly than the rest of the asset map.

What the chart is and is not saying

Bitcoin's monthly RSI at 2022 lows is a momentum statement, not a valuation statement. RSI is bounded by construction, and a reading near historic lows in a multi-month timeframe is consistent with sustained selling pressure but does not, on its own, identify a turning point. In November 2022 the RSI was printing at a similar level six weeks before the eventual bottom; in March 2020 it was printing near a level that marked a generational entry. The indicator is best read as a confirmation tool for a thesis already formed on other evidence, not as a thesis of its own.

The honest limitation on the public data set right now is that the sources do not break the $10 billion outflow by issuer, by jurisdiction, or by chain. Tether (USDT) and Circle (USDC) dominate the float, but the composition of the contraction has not been published. Cointelegraph's update also does not specify whether the move is net redemption into US dollars or rotation from stablecoins into spot bitcoin and ether. Both are plausible. They have different implications for downstream liquidity.

Stakes and the watchlist

If the stablecoin float continues to contract into the third quarter, the immediate effect is a thinner order book on centralised venues and a higher premium on collateral held against derivatives positions. The longer effect is harder to call. A slower-growing stablecoin market does not break the crypto thesis; it changes its funding profile. The market that emerges from this period will be one funded more by spot balances, by exchange-traded products, and by regulated US banks, and less by the synthetic-dollar float that defined the last cycle.

The single number to watch is the next monthly print of stablecoin total market capitalisation on the major aggregators. If the contraction slows in August, the rotation thesis weakens and the bid returns. If it accelerates, the question shifts from when the bid comes back to whether the new structural capex gravity is permanent.

This article builds on market reports circulated by Cointelegraph via Telegram on 11 and 12 July 2026. Monexus reads the stablecoin-outflow figure alongside the AI-infrastructure capex series and Bitcoin's monthly RSI print as a single liquidity story; the wire framing tends to treat them as separate desks.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://t.me/cointelegraph/293028
• https://t.me/cointelegraph/292910
• https://t.me/cointelegraph/292947
• https://t.me/cointelegraph/293028