I've spent the last 3 months building a personal system for reading macro + technical signals on ETH. I combine data analysis with AI (Claude Opus) to stress-test my reasoning, not replace it. The AI flags what I miss. This post is me following my own advice out loud.
This is not a prediction. It's a calculation with explicit assumptions.
The 90-day context
ETH is +8.78% over 90 days. 20 up days vs 12 down days. Price is currently ~$2,326, just above the EMA200 ($2,314) but below EMA20 ($2,338) and EMA50 ($2,354). That's a post-correction base-building regime, not a breakdown.
The March Iran-war drawdown took ETH to $1,972. That was the capitulation low. A ceasefire-driven rally followed, peaking at $2,434 on Apr 17. The current pullback is the third higher low in the recovery sequence. Structure is intact.
The ETH/BTC ratio (0.0308) is bouncing off 2026 lows. Average 90-day sentiment: 46 (Hopeful). Macro sentiment index: -12. The market absorbed the geopolitical shock and is rotating back toward risk. This matches the data.
The tactical setup
Entry zone: $2,260–2,320
Signals I'm watching:
- MACD bullish cross forming (histogram: +0.081)
- Bollinger Band squeeze at 2.54% width compression before expansion
- Funding rate negative (-0.000243) shorts are paying longs, squeeze setup
- Fear & Greed at 27 (Fear zone) historically favorable entry territory
- RSI14 at 38 oversold leaning, not extreme
The Apr 7–13 analog: post-spike compression resolved +7.5% higher. The current structure rhymes with that. The AI model flags a Day 3 shakeout toward ~$2,270 — which is actually my preferred entry, not a warning.
My 7-day forecast shows +6.8% expected move from the current level. I'm sharing this not as gospel but because the signal alignment between the model's triggers and what I observe technically is unusually strong. Confidence of 68% means the model itself is not certain and I'm not treating it as certainty either.
I ran this against war news, macro events, and funding data for the full 90-day window. The geopolitical noise has been priced in. The setup looks clean.
Where I'm wrong if this fails
- A close below $2,200 invalidates the higher-low structure
- New geopolitical escalation not in the 90-day data window
- BTC dominance continuing to rise rather than rotating into alts
- Funding squeeze resolves down instead of up (rare but possible)
Not financial advice. I'm posting this to think out loud and get pushback. If you see a flaw in the logic, say it plainly.
Thank you for reading.
I built a 90-day ETH model using data + AI. It just hit 68% BUY with 5/5 window agreement. Here's the full calculation.
byu/PeacefulNA inCryptoMarkets
Posted by PeacefulNA
1 Comment
Solid writeup, the invalidation criteria at the end is what separates this from most posts in this sub.
One thing worth stress-testing: the Apr 7-13 analog assumes similar liquidity conditions, but funding at -0.000243 is relatively mild negative — not the kind of extreme that historically precedes violent short squeezes.
The squeeze setup is there but the fuel might be limited.
The ETH/BTC ratio bouncing off lows is the signal I’d weight most heavily here. That ratio has been the leading indicator for alt rotation in every cycle, if it holds and turns, the technicals follow. If BTC dominance keeps climbing despite the macro rotation narrative, that’s your earliest warning before price confirms.
68% confidence with explicit assumptions and clear invalidation levels is more intellectually honest than most trading content online.
What’s the model’s historical accuracy on similar 5/5 window agreement setups, have you been able to backtest that specific confluence?