Glassnode adds proximity-weighted cost basis signals for BTC and ETH
News
Glassnode outlined a proximity-weighted cost-basis framework that turns CBD into support-resistance signals and proximity versions of MVRV, NUPL and premium.
Glassnode has published a new on-chain framework that turns Cost Basis Distributions into compact support and resistance signals by weighting nearby supply more heavily than distant supply. The October 20, 2026 methodology is intended to show where tradeable supply sits relative to spot, and to restate familiar valuation tools such as MVRV and NUPL on a proximity basis, according to Glassnode.
Evidence and context
In the post, authored by Vicky Yang, co-founder at Fountainhead Digital and published by Glassnode on Tue, 20 Oct 2026, the firm argues that traditional cost-basis measures are useful but incomplete because they can underweight nearby supply clusters that price is most likely to test. The proposed Proximity Support–Resistance framework, or PSR, addresses that gap by assigning exponentially decaying weights to 100 cost-basis quantiles according to distance from spot, with the decay scaled by the 30-day rolling standard deviation of price, the article says.
Under the method, proximity support is the weighted supply below spot and proximity resistance is the weighted supply above spot. Glassnode defines the Proximity Support Index as PSI = PS / (PS + PR) and the Proximity Resistance Index as PRI = PR / (PS + PR), which means the two series sum to 1. The background material supplied with the article says PSI above 0.5 indicates more proximity-weighted supply below spot than above it, while PSI below 0.5 indicates the opposite. The same proximity weighting is then used to compute a proximity realized price, from which proximity MVRV, proximity NUPL and Proximity Premium are derived.
The article presents historical examples in which shifts in PSI preceded changes in trend. Glassnode says BTC moved from roughly 29,000 to 26,000 between late July and mid-September 2023 while PSI remained mostly below 50%, then PSI stabilized above 50% in October before price advanced into the low 40,000s by December. It cites another example from February to April 2025 in which PSI moved from the high-30% area to the high-50% area while BTC traded near similar price levels, before a later rally. On the downside, the article points to late 2025 into January 2026, when BTC revisited about 85,500 on Dec. 18 and 86,600 on Jan. 25 while PSI fell from 57.5% to 48.7%, followed by a decline toward 65,000. It also describes a separate deterioration in PSI from the high-60% area to the low-40% area between mid-April and late May 2026, associated with a drop into the low-60,000s.
Glassnode also says it tested the approach in regressions for BTC and ETH across multiple forward-return horizons, comparing models that included recent price momentum alone with models that added one on-chain metric at a time. According to the article, the proximity-based measures added more explanatory power than short-term holder aggregates across horizons and both assets, with Proximity Premium described as statistically significant at the 1% level for both BTC and ETH. The post does not provide the full regression tables in the supplied text, so the underlying sample design and coefficients are not available here.
The methodology is designed for the current market structure, Glassnode says, because deeper derivatives markets and a larger base of synthetic exposure can make price action more two-sided and extend consolidation phases within broader trends. The firm frames PSR and the related proximity valuation measures as tools for identifying where nearby supply may act as support or overhead resistance, but the material supplied with the article describes them as signals, not guarantees of trading outcomes.
