US Spot ETF flows have swung back to persistent net outflows, with the 7-day average now near the weakest levels of the cycle.

The shift in flows has coincided with $BTC retracing from $82K to $69K, highlighting materially softer demand.
📉https://glassno.de/3POhzGl

At $69.5k, LTH Relative Unrealized Loss sits at 15.5%. For every dollar long-term holders' bags are worth today, they are carrying roughly 15 cents in unrealized loss. At cyclical extremes, that number has exceeded 50 cents on the dollar. Stress is present, but the long-term holder base remains far from the levels of pain that have historically marked cycle lows.

📉 glassno.de/lth-url-loss

Small caps surged in 2024. Yet raw aggregate market-cap data captured little of the rally.

The reason is structural: when an asset performs well and crosses the $100M threshold, it leaves the small-cap cohort taking its gains with it. The raw series systematically understates small-cap market moves.

As assets appreciate, decline, launch, and disappear, aggregate market analysis beyond price is increasingly difficult.

We just launched Glassnode Global Metrics: a new suite of aggregate metrics and normalized indices for consistent analysis of market-cap cohorts, sectors, and custom asset baskets through time.

Available via API.

Learn more → glassno.de/4uG6dmN
Explore the metrics → glassno.de/3RzjwqN

Prediction markets have a choke point: the API.

Every Polymarket and Kalshi order passes through a centralized endpoint, making network latency a direct input into order placement speed.

Where do you deploy from if you trade both venues? Polymarket pulls you towards London, Kalshi toward Ohio. You can't optimize for both.

Our live Latency Monitor for prediction market exchanges tracks real-time latency from probes deployed around the world → glassno.de/4o7fWAi

Access trusted digital asset data directly in your warehouse. Glassnode is now available on Snowflake.

▷ Query the full depth of our data in SQL like any other table in your warehouse. No custom pipeline to build, no schema to maintain, no ingestion lag.

▷ Access on-chain, derivatives, ETF, treasury, and spot market data in one place.

▷ Point-in-Time datasets for backtesting without look-ahead bias. Glassnode is the first provider to offer PiT blockchain data on Snowflake, because backtesting on retroactively revised data isn't backtesting. It's overfitting.

Nine years of proprietary methodologies, integrated seamlessly into the workflows you already run. Available as an add-on to the Professional plan.

🔍 Learn more → glassno.de/49F326D

❄️ Request access → glassno.de/4uMNKFd

The share of ETH supply sitting at more than 3x profit has dropped to 11%, the lowest reading since February 2017.

What makes this cycle structurally different?

In the previous two cycles, this cohort exceeded 50% of total supply at peak. This time, that threshold was never reached. ETH's profitability profile has fundamentally compressed relative to prior cycles.

📉 http://glassno.de/eth-supply-pnl-relative

The 90D-SMA of total fees paid on the XRP network has fallen from 5.9k XRP in Feb 2025 to 0.5k XRP today, a 91.5% decline.
A drop of this magnitude is not a fee market adjustment. It reflects a near-total contraction in organic transaction demand on the network since the speculative peak.

📉 glassno.de/xrp-fees

Nearly half of the $BTC circulating supply was in profit at the cycle peak. Today, that figure has fallen sharply as over 8M BTC sit underwater, highlighting the scale of the recent market reset.
📉https://glassno.de/4unfmj6

The 90D-SMA of XRP's Realized Profit to Loss Ratio has fallen to 0.38. For every dollar of loss being realized in the market, only 38 cents of profit is being taken.
At the 2025 peak, this ratio reached 50, meaning profit-takers were overwhelming loss-sellers by a factor of 50x. That dynamic has fully inverted.

A ratio this deep below 1 reflects a market where the majority of participants who are moving coins are doing so at a loss, a hallmark of intense capitulation.

📉 glassno.de/xrp-rplr