Key points: 

• Bond yields have been going up since the beginning of the Iran war, widely attributed to inflation expectations due to energy prices
• However, the five-year inflation expectation priced into Treasury Inflation-Protected Securities is 2.2% and trending down since May
• The driver appears to be rising real yields with bearish implications for yield-free assets like Bitcoin 

Continuation of Q2 bond selling

After yields reached local lows in early March, US government debt has been undergoing a multi-month sell-off. This week, after the most recent meeting of the Federal Open Market Committee (FOMC), 30-year Treasury yields made headlines by reaching the highest level since 2007. 

In line with the two-year yield rising 76 basis points (bps) in this window, a September rate hike by the Federal Reserve is priced into the markets at 63%, according to CME FedWatch.

2Y, 10Y and 30Y US Treasury Yields. Data Source: Treasury.gov

With rates at these elevated levels, government bond investments are, for the first time since 2019, more profitable than cash-and-carry trades in the crypto markets, as per Glassnode’s latest research. 

2Y US Treasury yield and crypto futures carry trade. Source: Glassnode

The mainstream inflation narrative

The reason for the bond sell-off is commonly taken to be the inflationary pressures from higher commodity and energy prices. The multi-month bond sell-off coincides with the start of the Iran war and resulting closure of the Strait of Hormuz. Furthermore, the daily closes of the two-year US government bond yield, West Texas Intermediate (WTI) and Brent Crude have correlated since March at a coefficient of r=0.44:

Daily closes of WTI and Brent Crude against 2Y Yield. Data Sources: fred.stlouisfed.org, EIA

WTI briefly rose once again above $85 a barrel on Thursday after President Donald Trump threatened Iran and bonds sold off leading into the FOMC. Nothing about the conflict suggests a near-term resolution, which has led some to argue that higher rates are being caused by inflation expectations.

WTI (West Texas Intermediate) oil price chart. Source: Tradingeconomics.com

This has driven loud inflation scares through the mainstream financial press, with recent Bloomberg headlines, such as “Global Bonds Are Reeling as Oil Surge Rekindles Inflation Threat”, “US Yields Hit Two-Month High as Oil Sparks Inflation Risk” or “Global Bond Selloff Worsens as Rising Oil Prices Spook Investors”. Among the ever-inflation-aware crypto and precious metals audience, this narrative is popular, too: 

By comparing the yield of a TIPS with a regular, equally dated Treasury, the expectation of future CPI inflation can be estimated as the so-called breakeven rate. And although Treasury yields have been rising, the five-year breakeven rate has gone down sharply since May. 

Five-year breakeven inflation rate. Source: fred.stlouisfed.org

At roughly 2.2%, the five-year breakeven expects the Fed to achieve its 2% target in the medium term. However, more telling is that the breakeven rate has been moving in the opposite direction to the nominal treasury yields.

While the five-year nominal yield rose 33 bps, TIPS data suggests this was the result of an 84 bps rise in the real yield, partially offset by a 51 bps decline in expected inflation. While the inflation narrative remains a compelling story, the marketplace says otherwise. The real story ought to be a rise in real yields. 

What it may mean for crypto

Generally, rising “real” investment returns on bonds and stocks in terms of CPI make non-yielding assets such as Bitcoin relatively less attractive to certain investors. Beyond this, the impact on the crypto market depends on the explanation for higher real rates, of which several are available. 

Reserve liquidation — No clear impact on Crypto. Higher oil prices widen trade deficits for Asian energy importers. As oil is generally priced and settled in US dollars, shortages in the local eurodollar markets in Asia have occurred, which has put their exchange rates under pressure. The Japanese yen (JPY), Philippine peso (PHP) and Indian rupee (RBI) have all required central bank intervention to defend their exchange rates. As these measures are funded by the sale of US Treasury reserves, this puts upward pressure on bond yields. HSBC’s Frederic Neumann is on record attributing the bond sell-off to FX pressure rather than a verdict on the dollar.

Demand destruction — Bearish for Crypto. An oil shock that persists long enough stops being inflationary and starts triggering a recession. Neuberger Berman argued in its second-quarter outlook that investors are underpricing the hit to output from sustained energy prices. The credit contraction that coincides with a recession would be bad for equities and Bitcoin by severely restricting liquidity. In a real sign of recessionary credit events, credit spreads are expected to widen. Cointelegraph reported on possible first signs of this on Wednesday.

Related: Cost to insure AI debt reaches record high amid Asian semiconductor tumble

Investment demand — Likely bearish for Crypto. Real rates may have also responded to expected growth and the demand for capital from the AI sector. Government bond issuance is increasingly competing with the record issuance of corporate bonds from AI hyperscalers. Goldman Sachs Research projects roughly $755 billion of AI capex in 2026 and about $920 billion in 2027. UBS has raised its 2026 investment-grade issuance forecast to $1.8 trillion, with technology supply lifted to $360 billion on hyperscaler guidance. As crypto is competing for a similar pool of capital and investor cohort, this is likely to suppress the sector.

A Bank of Italy study found that stablecoin-based remittances did not offer a systematic cost or speed advantage over traditional payment channels, as fiat on- and off-ramp frictions accounted for most costs and transfer delays.

Researchers tested 200 USDC (USDC) remittances across 10 bidirectional payment corridors linking Italy with Brazil, Argentina, Japan, the United Arab Emirates and South Africa, comparing end-to-end costs and settlement times with traditional remittance services. They found that exchange fees and currency conversion made up most of the cost, while blockchain transaction fees represented only a small share.

Geographic design of the remittance experiment. Source: Bank of Italy

Across the stablecoin remittances examined, total costs ranged from 0.3% to nearly 9% depending on the payment corridor, while transfers settled in less than 20 minutes where instant payment systems were available and one to two business days where they were not.

Using the World Bank’s reported global average remittance cost of 6.65% as a benchmark, the study found stablecoin transfers were cheaper in most of the payment corridors examined. However, they were less expensive than Wise in only three of seven comparable corridors.

Related: Europe should weigh tokenized SEPA payments, Bank of Italy official says

Payment infrastructure remains critical

The study concluded that investment in domestic instant payment infrastructure could improve the competitiveness of stablecoin-based cross-border payments, finding that settlement times depended heavily on the quality of local payment rails. 

The authors argued that the biggest gains may come when stablecoins no longer require conversion back into fiat currency, writing:

If stablecoins could be spent directly in the real economy, for goods and services, rents, or school fees, without reconversion into local fiat currency, the economic advantages of stablecoin-based transfers would be substantially higher.

Regulation shapes remittance efficiency

The study also found that regulatory design played a major role in determining transfer efficiency. The authors said prohibitionist regulatory regimes failed to fully suppress stablecoin demand and instead pushed users toward offshore platforms and other unregulated channels, while overly restrictive frameworks increased operational complexity for retail users. 

The findings come as the European Union has implemented its Markets in Crypto-Assets (MiCA) framework and the United States has enacted the GENIUS Act, two regulatory regimes that govern crypto assets and payment stablecoins, respectively.

The stablecoin market has grown to about $307 billion, up roughly 16% over the past year, according to DefiLlama data.

Magazine: A quantum roadmap would push Bitcoin much higher: Charles Edwards

A proposed Aave governance initiative would wind down multiple Aave V3 lending deployments on six blockchains and retire a large set of low-usage token listings. The plan, advanced through the protocol’s ARFC process, targets a cleanup covering $98.1 million in supplied assets and $15.6 million in outstanding debt, based on balances recorded on July 28.

According to LlamaRisk, which worked with Aave service providers on the assessment, the proposal recommends offboarding 50 low-use reserves and retiring 21 matured Pendle principal token listings across 11 deployments. It also calls for retiring all 25 reserves on Sonic, Scroll, zkSync, Metis, Soneium, and Aptos.

Key takeaways

• The ARFC would deprecate Aave V3 markets on Sonic, Scroll, zkSync, Metis, Soneium, and Aptos, alongside removing 50 low-use reserves and 21 matured Pendle principal token listings.
• The scope is tied to on-chain balances measured July 28, with $98.1 million supplied and $15.6 million in debt included in the cleanup.
• Risk service provider LlamaRisk characterizes the action as part of Aave’s broader risk-governance frameworks rather than a reversal of its multichain growth thesis.
• Prior multichain “temp check” voting already shut down underperforming instances on zkSync, Metis, and Soneium and set a $2 million annual revenue floor for new deployments.
• Aave founder Stani Kulechov framed the move as reducing both economic and technical risk surface under updated listing and risk frameworks.

What the ARFC would change in Aave V3

An ARFC—an “Aave Request for Comment”—is presented as a detailed governance proposal and precursor to an Aave Improvement Proposal. It is not itself confirmation that final on-chain voting has been completed or that execution is already underway.

In this case, the recommendation focuses on reducing exposure to markets with limited usage or maturing positions. LlamaRisk’s work with other Aave service providers outlines multiple categories of deprecation: low-use reserves and certain Pendle principal token listings that have matured, alongside full reserve retirements on the six named chains.

Aptos exit arrives after a rapid liquidity decline

The inclusion of Aptos stands out because it follows relatively recent deployment activity. LlamaRisk’s materials indicate Aave launched its V3 market on Aptos about 11 months earlier. In that period, liquidity fell by 94% over six months, and quarterly revenue reportedly dropped below $1,000, according to LlamaRisk.

Under the proposal, not all chains are treated the same way. LlamaRisk states that every reserve on Scroll, zkSync, Metis, and Soneium was already frozen. By contrast, Sonic and Aptos remained active at the time of the snapshot, with the ARFC recommending that they be frozen as well.

This structure matters for how quickly deprecations could translate into actual risk reduction. Freezing already stops new activity, but full retirement would further narrow Aave’s operational footprint on those deployments.

How earlier “temp check” decisions set the stage

The ARFC is not the first governance signal that Aave would be willing to scale back underperforming V3 instances on certain chains. A prior “temp check” on Aave’s multichain strategy concluded on Dec. 5, 2025, according to the governance record referenced in the source materials. That vote reportedly returned 923,400 votes in favor and under 1% against changing reserve behavior for underperforming instances.

That earlier governance outcome included actions affecting zkSync, Metis, and Soneium—specifically shutting down instances—and introduced a $2 million annual revenue floor for new instance deployment. In other words, the latest ARFC reads less like a sudden pivot and more like an operational follow-through on criteria that were already accepted by the community.

Risk framework updates and protocol-wide cleanup logic

The proposal aligns with Aave’s evolving risk and listing governance. The source notes that Aave added Scroll to the affected set through an accelerated process in April, referencing a direct-to-AIP proposal. In that description, the measure was framed as completing Scroll’s deprecation after a rapid deterioration in network liquidity and Aave market activity.

Separately, Aave published an updated risk framework on June 9 covering asset, bridge, monitoring, and chain risk, along with criteria for winding down reserves or deployments. The current ARFC announcement, as described in the source materials, suggests “de facto” adoption of these rules for the present cleanup.

Aave founder Stani Kulechov also addressed the initiative in a Thursday social media post. He said the move would “reduce Aave’s economic and technical risk surface as part of the new Aave Risk Framework and Technical Asset Listing Framework.” He further emphasized that the action “is not a reversal” of the protocol’s multichain expansion strategy, and that Aave would continue continuous risk assessment across deployments.

That distinction is likely important for market participants. Aave’s multichain approach appears to remain intact conceptually, but the governance direction points toward tighter enforcement of performance and risk thresholds—essentially focusing capital and attention on deployments that meet criteria and exiting those that do not.

Why this matters for users and market participants

For users and liquidity providers, deprecations can change the path of capital: liquidity may diminish further as reserves are frozen or retired, and markets tied to low-use reserves can become less accessible over time. For borrowers and lenders, winding down V3 markets can also affect how easily positions can be adjusted, particularly if token listings tied to specific assets or principal tokens are retired after maturity.

For investors and governance observers, the bigger signal is how Aave is operationalizing its frameworks. By connecting deprecations to measurable liquidity and revenue outcomes—and by referencing an earlier temp check that set a revenue floor—the ARFC underscores a governance style that is increasingly rules-driven rather than ad hoc.

Readers should watch for the next procedural steps: whether the ARFC proceeds into an Aave Improvement Proposal for formal voting, and how execution is sequenced across the chains involved—especially where Sonic and Aptos were still active at the time of the July 28 snapshot.

The exchange hit record market share and crossed $100 million in prediction market revenue, but a $1.36 per share loss and declining transaction fees show the Everything Exchange still runs on a shrinking engine.

Coinbase delivered its third consecutive quarter of missed revenue estimates on July 30, posting a $359.5 million net loss that turned a year of strategic diversification into a question about whether any amount of product expansion can offset a sustained decline in trading fees.

The headline numbers were difficult to frame positively. Revenue of $1.22 billion missed consensus by roughly $80 million. Earnings per share came in at negative $1.36, far below estimates that ranged from negative $0.01 to positive $0.14 depending on the source. Adjusted EBITDA of $207.8 million missed by 31%. The stock dropped more than 5% in after hours trading before partially recovering the following day.

Yet beneath the miss, the structural story is changing. Bitcoin now accounts for just 12% of total revenue, down from more than 50% historically. Subscription and services revenue has grown from $6 million per quarter in 2020 to $555 million today. Prediction markets crossed $100 million in annualized revenue. The company Brian Armstrong calls the Everything Exchange is genuinely becoming one. The question is whether it is becoming one fast enough to survive the quarters when its original business contracts.

The revenue miss, decomposed

The $80 million revenue gap was not concentrated in a single segment. It was distributed across nearly every line item, suggesting the problem was market wide conditions, not a specific operational failure.

Transaction revenue came in at $599 million against an estimate of $640 million. Consumer trading, still the largest single revenue source at $452 million, missed by $40 million and declined 30.5% year over year. The total crypto market capitalization fell 11% during the quarter, and spot trading volumes dropped 25%. Coinbase was swimming against a current that pulled the entire industry down.

Institutional trading was the exception. Revenue of $100 million beat estimates of $116 million in absolute terms but represented a 64.6% year over year increase. Coinbase is gaining institutional share even in a declining volume environment, a pattern that suggests its expansion into tokenized stocks and international markets is generating durable demand instead of speculative volume.

Subscription and services revenue of $555 million missed its $601 million estimate by roughly $45 million. Within that category, stablecoin revenue of $292 million came in below the $339 million consensus, down 12.1% year over year. The $47 million stablecoin miss was the largest single line item shortfall in the subscription segment and represents the first time USDC revenue has disappointed at this scale since the revenue sharing arrangement with Circle began generating material income.

Blockchain revenue of $83 million missed by $13 million and declined 42.3% year over year, reflecting lower activity on Base chain during the broader market cooldown. Other transaction revenue of $47 million also came in light at $53 million estimated. Only interest and finance fees, at $66 million, beat estimates, rising 11.5% year over year. The interest income beat is a direct consequence of elevated USDC balances earning yield in a high rate environment, a tailwind that could reverse if the Federal Reserve begins cutting rates.