📖 How to Read the Bayesian Stock Forecast Terminal

This page explains how to read the terminal and how each forecast is calculated. The goal is to give you the concepts in plain language — not the internal code structure — so you can interpret any ticker run with confidence.

Before evaluating any trade setup, look at the Validation block. This is the ultimate guardrail for your capital.

• What it does: The terminal uses expanding-window walk-forward validation. It steps through history one period at a time, makes out-of-sample predictions, and scores them against "climatology" — the historical base rates of the asset — using log-loss and accuracy.
• The Blend Rule: When the network's historical predictions are uncertain, the forecast is automatically shrunk toward the historical prior. The blend weight is chosen by whichever mix performed best in the walk-forward test.

How to read it:

The terminal pulls daily open-high-low-close-volume data for the ticker and aligns it with the Nasdaq-100 (QQQ) as a broader market benchmark. Each indicator is then categorized into discrete states:

Using today's discrete state as evidence, the terminal runs exact Bayesian inference by enumeration. It outputs the mathematical probability of the asset gaining more than 10% or losing more than 10% over 15-trading-day (3 weeks) and 20-trading-day (4 weeks) horizons.

The terminal does not treat every indicator equally. It ranks all six technical indicator nodes by Mutual Information (MI) against the forward-return target bucket.

It then selects only the top 3 highest-ranking nodes to drive the inference model, fitting conditional probability tables with Laplace/BDeu smoothing. The indicator with the highest MI score is reported as the primary market driver for that ticker right now.

The terminal isolates every historical day in the asset's history that shared the same combination of discrete states as today. These are the historical analogues.

• Sample Count ("Cell n"): When the current state cell has ≥20 historical samples, the median target and 80% band come from the empirical forward-return distribution of those exact matching days. If it has fewer than 20 samples, it safely falls back to a lognormal distribution band at current realized volatility.
• The Forward Path: Use the 20-day cumulative median return array to optimize trade entry timing. If Days 1–5 show a negative path trend, history suggests waiting for an initial shakeout or pullback before entering.

• Support & Resistance: The terminal builds a price ladder using swing pivots and volume-by-price nodes from the last year of trading data, assigning each level a strength score.
• First-Touch Barriers: If you input an optional Entry, Stop-Loss, and Target on the run screen, the terminal executes a continuous Geometric Brownian Motion (GBM) Monte Carlo simulation. It reports the probability of your target being hit before your stop, plus the mathematical Expected Value (EV) of the trade.

Always remember the boundaries of pure mathematical models:

1. 1. Technical Boundary: The model only sees price action, volume metrics, and index states. It cannot see unannounced macro shocks, earnings reports, corporate restructuring, or regulatory shifts.
2. 2. The Context Filter: Use the optional News Overlay to bridge this gap. If the math shows a strong technical edge, but live headlines indicate severe fundamental overvaluation or catalyst risk, adapt your trade thesis accordingly.

Not investment advice.

These are the features we're considering. Click "Want this" on anything you'd actually use — your votes decide what we build first.

One vote per feature per person. Results are public on this page.