Originally published at satoshimacro.com.
When I built the SatoshiMacro Model for Bitcoin, the obvious next step looked easy: swap the ticker and ship an Ethereum version. It was not easy, and the reasons why say a lot about how differently these two assets actually behave.
SMM-ETH is the Ethereum variant of my six-tier cycle confluence model. Same architecture as the Bitcoin version: weighted tiers covering cycle timing, valuation, sentiment, rotation, on-chain economics and macro, each signal expressed as a lookahead-free expanding-window percentile so historical readings never peek at future data. But almost every tier needed rethinking.
Start with cycle timing. The Bitcoin model leans on the four-year halving cycle. Ethereum has no halving. Post-merge, there is no scheduled supply event to anchor a clock to, so the ETH timing tier drops the halving inputs entirely rather than pretending an analogue exists.
Valuation needed its own regression. Bitcoin's power law fits with one slope; Ethereum's price history fits a much shallower power-law slope of roughly 2.3, with its own moving-average structure. Forcing Bitcoin's curve onto Ethereum produces readings that are confidently wrong, which is worse than no reading.
The tier that makes SMM-ETH genuinely different is rotation. Ethereum does not just have a cycle; it has a position inside crypto's internal rotation. The ETH/BTC ratio percentile, Bitcoin dominance (inverted) and the Altcoin Season Index anchor this tier, because late-cycle behaviour for ETH is historically about capital rotating down the risk curve, not just prices going up.
Two honesty notes I refuse to bury. First, the sentiment tier currently uses several Bitcoin-side inputs (funding, open interest, Coinbase premium) as broad-crypto risk-on proxies, and the page labels them exactly that way, pending dedicated ETH feeds. Second, the model is five tiers live out of six: the validator/staking tier is still in development, because Ethereum's on-chain economics run on validator behaviour, not miner behaviour, and copying the Bitcoin miner inputs across would be fake precision.
Calibration is honest too: across the five major ETH inflections since 2018, the model lands four in their target zones. The miss is the 2022 bottom, which reads Neutral rather than Accumulation, and my working hypothesis is that the missing validator-stress dimension is exactly what would have caught it.
The full model, methodology, and every signal's current reading are free at https://satoshimacro.com/tools/crypto/satoshimacro-model-eth/ - and the Bitcoin original is at https://satoshimacro.com/tools/crypto/satoshimacro-model/ if you want to compare the two panels side by side.
Disclosure: I built and maintain SatoshiMacro. The model is free and ad-supported (broker affiliate links on the main site, not in this post). This post is editorial, not financial advice.
