Bitwise finds Bitcoin dictates Ethereum price, fundamentals show minimal influence
Bitwise: ETH still trades as BTC beta. A new model shows BTC drives most weekly ETH moves since 2018.
Ethereum “moves nearly 1:1 with” Bitcoin on a weekly basis. Bitwise estimates a BTC coefficient near 0.99 and roughly 65% of ETH return variance explained by BTC alone (Bitwise factor model).
“Adoption fundamentals… have less impact on Ethereum’s price than many assume.” Bitwise adds that revenue was dropped as “noise rather than signal,” framing ETH “more like a network-driven commodity than a business with durable cash flows” (Bitwise factor model).
What else matters, and when:
- Financial conditions measured by the Bloomberg US Financial Conditions Index had a coefficient near 0.05. Mean explanatory power 11.3%, peaking around 40% (Bitwise factor model).
- Network activity proxied by active addresses showed a coefficient near 0.03. Average 6% explanatory power, up to 30% in stronger phases (Bitwise factor model).
- ETF flows carried a small coefficient near 0.01 but were highly significant. About 10% on average, up to 40% at peaks (Bitwise factor model).
Regimes shift the betas. From June to August 2025, ETH behaved like levered BTC with a BTC coefficient around 1.5–1.6 as BTC neared highs (Bitwise factor model).
In H2 2022 after FTX, BTC explained up to 90% of ETH returns. Other factors flipped negative as liquidity dominated, and ETH was “anchored to BTC” (Bitwise factor model; context on FTX collapse NewsBTC background).
There were exceptions. In May 2021, BTC sensitivity was lowest as NFTs pushed active addresses higher and ETH kept rallying (Bitwise factor model).
Forecasting edge stayed simple. Out-of-sample, the multi-factor GETS model did not beat a basic AR(1)+BTC approach. Most predictive value came from BTC exposure and price persistence, with limited extra lift from other variables (Bitwise factor model).
Bitwise calls this a “paradoxical” setup. Institutional relevance and tokenization strength grow, but price remains driven mainly by external beta (Bitwise factor model).
