ETH Security Mechanism
Stablecoin TVL secured on Ethereum vs. ETH market cap — and what the ratio implies about fair value. Model R² = 0.86 on data since 2020.
What this shows
As of March 2026, Ethereum held ~$164B in stablecoins — within 3% of an all-time high — while ETH's market cap sat at ~$239B, down 59% from peak. The TVL/Mcap ratio at 0.69 was at the 99th percentile of all historical readings.
Every time this ratio has approached 0.78–0.80, one of two things has happened: ETH price has recovered, or stablecoin growth has stalled. It has never sustainably broken above that ceiling.
Read the full analysis
Why fees stopped being the right signal after EIP-4844. The institutional shift in stablecoin composition. What the model's current Z-score of −1.14 implies for positioning.
Read on Substack ↗ View editorial with full chart →ETH Combined TVL Model
Combined stablecoin, DeFi, and RWA TVL secured on Ethereum vs. ETH market cap. Model R² = 0.902. Current Z-score: −0.97σ below fair value.
What this shows
The updated model expands beyond stablecoins to include DeFi TVL ($46B) and RWA tokenization ($17B) — both of which create genuine security dependencies on Ethereum. Combined TVL now sits at $230B, approaching the all-time high of $271B set in late 2025.
The R² improves from 0.86 (stablecoins only) to 0.902, capturing a more complete picture of the economic activity Ethereum's security budget is asked to protect. At a Z-score of −0.97, ETH is roughly one standard deviation below model fair value.
Read the full analysis
Why DeFi and RWA create independent security dependencies. The RWA growth curve and what it means for model price over the next two years. What a Z-score of −0.97 has historically implied.
Read on Substack ↗ View full analysis →ETH Stacked TVL
ETH price against stacked Stablecoin, DeFi, and RWA TVL — showing rolling all-time highs and how total secured value has evolved across cycles.
What this shows
Total value secured on Ethereum broken out by category — stablecoins, DeFi TVL, and RWA tokenization — stacked against ETH price on a log scale. The rolling all-time high line shows how close total secured is to its prior peak of $271B (late 2025).
The divergence between ETH price and total secured TVL is the central tension of the current cycle: the base of value sitting on the network has recovered while ETH's market cap remains well below its own prior highs.
Read the full analysis
Why DeFi and RWA create independent security dependencies alongside stablecoins, and what $230B secured relative to a $239B market cap implies for ETH's price floor.
Read on Substack ↗ View full analysis →ETH Proxy Confluence Signal
Three independent conditions — RV7/RV30 vol spike, 90-day drawdown, and combined TVL model discount — that have historically converged at major ETH entry points.
What this shows
A confluence signal derived entirely from price and on-chain TVL data. The three conditions — short-term vol spiking above medium-term (RV7/RV30 > 1.7×), ETH at least 20% below its 90-day high, and ETH price at or below −1σ of the combined TVL regression model — must fire simultaneously for a signal.
Each condition captures a different dimension of stress: volatility regime, price momentum, and fundamental valuation. Confluence is rare by design and historically meaningful when it occurs.
Read the full analysis
How the three-condition framework was constructed, the forward return distribution at each signal tier, and what the combined TVL model implies about ETH fair value at current levels.
Read on Substack ↗ View full analysis →BTC Confluence Signal
Miner production cost proximity, RV7/RV30 volatility spike, and 90-day drawdown depth — three independent signals that have historically converged at major Bitcoin cycle lows.
What this shows
Green dots mark days when all three conditions align simultaneously: BTC price at or below 1.5× estimated miner breakeven cost, RV7/RV30 ratio above 1.8×, and a 90-day drawdown of at least 20%. Since 2014 this has fired at late 2018, March 2020, late 2022, and February 2026.
Each signal is drawn from a different data source and economic mechanism — making confluence rare by design, and historically meaningful when it occurs.
Read the full analysis
The construction of each signal, full model assumptions, hash rate growth dynamics, and what the February 2026 confluence implies for investors with an 18-month horizon.
Read on Substack ↗ View full analysis →MSTR P/NAV — Z-Score Heatmap
MicroStrategy's market cap premium to its Bitcoin NAV, shown as a z-score heatmap alongside BTC price and estimated miner production cost. Daily updated.
What this shows
MSTR's P/NAV ratio measures its market cap relative to the dollar value of its Bitcoin holdings. The z-score heatmap (calibrated to the post-2022 era) shows when the premium is historically extended or compressed. Red bars signal elevated premium risk; green bars signal compression toward fair value.
The BTC production cost overlay (blockchain.com difficulty data) provides context for BTC's own support floor — when BTC approaches production cost, the denominator of P/NAV itself is under pressure.
How to use it
A z-score above +2 has historically preceded periods of P/NAV compression. A z-score below −1 with BTC near production cost has marked combined stress events. The neutral zone (−1 to +1) is where MSTR spends most of its time post-2022.
Read on Substack ↗BTC Weekly Price & RSI(14)
BTC weekly price on a log scale with RSI(14) regime dots — four states: rare oversold recovering (green), rare oversold falling (amber), rare overbought running (gold), rare overbought topping (red). Daily updated.
What this shows
RSI(14) computed on weekly BTC closes. Each dot is colored by its percentile rank over full history: the bottom 12th percentile (rare oversold) and top 88th percentile (rare overbought) are the signal zones. Within each zone, direction of the 3-week RSI slope determines the sub-state.
Triangles on the price panel mark the same extreme weeks — up-triangles at rare oversold, down-triangles at rare overbought — making the historical cluster patterns visible against price.
Signal states
Green: rare OS + RSI rising (confirmed reversal). Amber: rare OS + RSI still falling (knife-catch risk). Gold: rare OB + RSI rising (extended momentum). Red: rare OB + RSI topping (rolling over). Steel: neutral mid-field.
Read on Substack ↗ETH Weekly Price & RSI(14)
ETH weekly price on a log scale with RSI(14) regime dots — the same four-state color scheme calibrated to ETH's full history back to 2017. Daily updated.
What this shows
ETH's RSI(14) on weekly closes, percentile-ranked against its full history from 2017. The rare oversold zone (≤12th pct) has historically marked major accumulation windows — 2018 bear, 2020 COVID crash, 2022 bear, and the 2025–26 drawdown. The rare overbought zone (≥88th pct) has marked cycle peaks and local tops.
ETH's RSI distribution is structurally different from BTC's due to its deeper and more prolonged bear cycles, making the percentile approach more meaningful than fixed 30/70 levels.
Signal states
Same four-state framework as BTC: green (OS recovering), amber (OS falling), gold (OB running), red (OB topping). The slope is computed as a 3-week RSI difference, smoothing out single-week noise while remaining responsive to trend changes.
Read on Substack ↗BTC Annual RSI(52) Cycle
RSI computed on a 52-week (1 year) lookback — a slower oscillator that tracks BTC's full market cycle rather than short-term momentum. Daily updated.
What this shows
RSI(52) captures BTC's cycle-level momentum rather than weekly noise. It has historically peaked in the upper 80s–90s at major cycle tops (2017, 2021) and troughed in the 30s–40s at cycle bottoms (2018–19, 2022, early 2024). The current reading tells you where in the macro cycle BTC sits relative to its own history.
The rare zone thresholds (≤12th pct, ≥88th pct) are recalibrated against RSI(52)'s own distribution, which is tighter and slower-moving than RSI(14).
Why 52 weeks
A 52-week RSI spans exactly one year of weekly data — aligning with BTC's 4-year halving cycle and making each reading a direct comparison of the current year vs. the prior year. It's slow enough to filter weekly noise, fast enough to capture cycle turns within 3–6 months of a top or bottom.
Read on Substack ↗ETH Annual RSI(52) Cycle
RSI(52) on ETH weekly closes — the cycle-level momentum oscillator showing where ETH sits in its macro cycle relative to its own history back to 2017. Daily updated.
What this shows
ETH's RSI(52) has a distinctive pattern: it spends long periods in the mid-40s to mid-50s range during consolidation phases, then compresses sharply below the rare OS threshold during prolonged bear markets. The 2022–23 bear kept RSI(52) in the rare OS zone for over a year — the longest such period in ETH's history.
The current reading of ~47.5 at the 12th percentile marks ETH as sitting at the rare OS threshold on the annual cycle — a level associated with prior major accumulation windows when the slope turns positive.
ETH vs BTC cycle dynamics
ETH's RSI(52) cycle is typically more extreme than BTC's in both directions — deeper OS troughs and higher OB peaks — reflecting its higher beta to risk appetite. This makes the annual RSI signal potentially more informative for ETH than for BTC at cycle extremes.
Read on Substack ↗ETH Price + TVL Momentum
ETH price alongside daily total value locked Δ% (stablecoins + DeFi + RWA) and 30-day cumulative TVL return. Updated daily via GitHub Actions.
What this shows
Three-panel view of ETH's price alongside the daily momentum of total value locked on Ethereum — combining stablecoin supply, DeFi TVL, and RWA TVL into a single aggregate. Panel 2 shows green/red bars for each day's net TVL Δ%, while Panel 3 shows the 30-day cumulative TVL return from a common base.
TVL momentum diverging from price momentum has historically preceded major re-ratings in both directions — periods where TVL holds up while price falls often mark accumulation zones.
Data sources
Stablecoin TVL from DeFiLlama's stablecoin charts API. DeFi TVL from DeFiLlama's Ethereum chain TVL endpoint. RWA TVL aggregated from all DeFiLlama-tracked RWA protocols with ≥$1M on Ethereum mainnet. ETH price from Yahoo Finance daily close.
Read on Substack ↗ETH TVL Daily Component Flows
Stablecoin, DeFi, and RWA contributions to total ETH TVL change — with ETH price and 30-day cumulative returns for both ETH and TVL. Past 30 days, daily updated.
What this shows
A three-panel decomposition of Ethereum's TVL: ETH price (top), stacked bars showing each component's daily contribution as a percent of total TVL (middle), and the 30-day cumulative return for both ETH and total TVL (bottom). The stacked bars isolate whether moves in aggregate TVL are driven by stablecoins, DeFi protocols, or RWA tokenization on any given day.
The divergence between ETH cumulative return and TVL cumulative return in Panel 3 is one of the clearest short-window signals of over- or under-valuation relative to on-chain fundamentals.
Reading the chart
Gold bars = stablecoin contribution. Grey bars = DeFi TVL contribution. Green bars = RWA TVL contribution. The white dotted line tracks the total TVL Δ% — the net sum of all three components. When ETH's cumulative return (gold line, Panel 3) runs well above TVL cumulative return (green dotted), the market is pricing in expected TVL growth that hasn't yet arrived.
Read on Substack ↗ETH Combined TVL Model Z-Score
Log-linear regression of ETH price against combined TVL (stablecoins + DeFi + RWA) with ±1σ and ±2σ bands. Bottom panel shows the rolling Z-score — how many standard deviations ETH trades above or below model fair value.
What this shows
A log-linear model regressing ETH price against the total value secured on Ethereum — stablecoins, DeFi TVL, and RWA tokenisation combined. The shaded bands show ±1σ and ±2σ around the model's fair value line. Below the chart, the Z-score expresses in standard deviations how far ETH currently trades from where the model expects it.
Historically, Z-scores below −1σ have coincided with strong forward returns; readings above +2σ have marked cycle tops or preceded significant drawdowns.
Reading the chart
Gold line = ETH price. White dashed line = model fair value. Z-score bars shade green when ETH is cheap relative to on-chain TVL and red when expensive. The 28-day MA smooths out daily noise to show the regime trend.
Read on Substack ↗ETH Stacked TVL + Price | TVL / Market Cap
Stacked area of stablecoin, DeFi, and RWA TVL secured on Ethereum (left axis) alongside ETH price on a log scale (right axis). Bottom panel shows the ratio of total secured TVL to ETH market cap with a 30-day moving average.
What this shows
The top panel stacks the three pillars of value secured on Ethereum — stablecoins (gold), DeFi TVL (grey), and RWA tokenisation (green) — and overlays ETH price on a log scale to show how price has tracked aggregate TVL growth over time. The bottom panel distils this into a single ratio: total secured TVL divided by ETH market cap.
When the TVL/Mcap ratio approaches historical highs, ETH is absorbing more economic value than its market cap reflects — a condition that has historically preceded price mean-reversion upward.
Reading the chart
Gold area = stablecoin TVL. Grey area = DeFi TVL. Green area = RWA TVL. Gold line on right axis = ETH price (log). Bottom panel gold line = 30-day MA of the TVL/Mcap ratio. Rising ratio with flat or falling price is the key divergence to watch.
Read on Substack ↗BTC Production Cost Momentum
BTC price alongside daily mining production cost Δ% and 30-day cumulative cost momentum. Production cost modelled from hashrate, efficiency schedule, electricity price, and halving-adjusted block rewards. Updated daily.
What this shows
Three-panel view tracking BTC price against the daily momentum of its estimated production cost — the break-even price per BTC mined given current hashrate, miner efficiency, electricity costs, and the halving schedule. Panel 2 shows day-over-day production cost Δ%, Panel 3 shows the 30-day cumulative shift.
Production cost acts as a gravitational floor in bear markets and a benchmark for margin compression at cycle tops. When BTC price drops toward or below production cost, miner capitulation risk rises — historically a high-conviction accumulation signal.
Model methodology
Hashrate sourced daily from mempool.space (blockchain.info as fallback). Efficiency schedule steps down with each hardware generation cycle. Electricity assumed at $0.05/kWh with 1.10× PUE overhead. Block reward adjusts at each halving date. The result is a daily bottom-up production cost estimate, not a smoothed proxy.
Read on Substack ↗