1–3 month research → trade plan

Indikators turns market data into a mid-term BUY / SELL / HOLD — with entry, limit sell, and stop.

Stocks get fundamentals, technicals, earnings, SEC filings, and headlines in one brief. A local LLM scores it for a 1–3 month horizon and publishes an actionable trade plan. History stays public.

Sign in with Google How a recommendation is made ↓ Browse live recommendations

01 — Pipeline

How a recommendation is made

The model never invents quotes, earnings, or filings. It only reasons over a structured 1–3 month research brief.

  1. 01

    Ingest

    Price, valuation, growth, technicals, volume, sector tape, VIX risk flag, earnings window, SEC filings, and headlines.

  2. 02

    Research brief

    Data is formatted into a mid-term equity brief (or intraday crypto brief) with an explicit analysis framework.

  3. 03

    Model score

    Local Gemma returns JSON: signal, confidence, 1–3 month horizon, thesis, risks, entry zone, limit sell, and stop.

  4. 04

    Publish & verify

    The call is stored with timestamp. Later we attach the real price path so history is auditable — not marketing copy.

02 — Two paths

Stocks and crypto are not scored the same way

Stocks · weeks to months

Comprehensive brief

  • FFundamentals — P/E, P/S, D/E, revenue & EPS growth, margins, ROE/ROA, dividend, analyst consensus / PT, sector & industry.
  • TTechnicals — trend from SMA stack, RSI(14), full MACD (line / signal / hist), Bollinger position, computed support & resistance.
  • VVolume & tape — current vs 20d average, volume trend, S&P 3m, sector leaders/laggards, recent headlines.

Crypto · hours to days

Candle structure

  • 1mRecent one-minute OHLCV for short-horizon momentum and volume spikes.
  • 5mAggregated five-minute bars set the primary trend; 1m confirms or denies.
  • S/RSwing highs/lows become support, resistance, stop, and take-profit — formation named only if clear.

03 — Discipline

Signal rules the model must follow

High confidence is earned. Mixed evidence defaults to HOLD — we would rather miss a trade than invent conviction.

BUY

Constructive trend plus acceptable valuation/growth, or a clear mean-reversion setup with a catalyst.

SELL

Broken trend or overextension with deteriorating fundamentals or a concrete downside catalyst.

HOLD

Conflicting evidence, range-bound price action, or no edge after reading the brief.

Confidence scale

  • ≥ 70 — fundamentals and technicals clearly align
  • ~ 40–69 — lean one way, but something important is missing or mixed
  • ≤ 55 on crypto — preferred when 1m and 5m timeframes disagree or the tape is choppy

Analysis

Two short paragraphs: fundamentals, then technicals — citing numbers from the brief (≈120–220 words).

Trade plan

Entry zone, limit sell / take-profit, and stop — with support & resistance for context.

Horizon

Stocks: 1–3 months. Crypto: hours to a few days. Matched to the thesis, not day-trading noise.

Risks

Top 2–3 concrete risks to the call — what would invalidate the thesis.

04 — Live book

Recent recommendations

Open any symbol for the full thesis. History is free; this week’s freshest call unlocks with Pro.

All symbols →

Stocks

Crypto

Verified outcomes

Past signals show the real price move weeks later — so you can audit the model, not trust a slogan.

Evidence over hype

The brief is numeric. The thesis must cite it. No invented earnings, no invented prices.

Pro for the latest

History stays public. This week’s call, confidence detail, and targets are on the Pro plan.