DG

What it is. Each cell is one trading session's 14-day Money Flow Index for DG, recorded from the daily OHLCV series — the same deterministic engine that scores the sector orbs. Colour follows a 10-step scale: red = net selling pressure, grey = neutral, green = net buying pressure.

How to read it. Runs of green mark sustained accumulation; runs of red mark sustained distribution; grey stretches are chop. The streak badge counts consecutive sessions on the same side of the 60 / 40 thresholds. Hover any cell for its exact date and value.

The record. A new cell is appended after every US close by the nightly data run, and the record keeps growing beyond the chart's window. The latest cell always equals the 14-Day MFI Score in the header above. See How the dial is built for the formula.

Honest labels. Technical money-flow context from price and volume — not dollar flows, not investment advice. See the disclaimer.

What it is. This page applies the exact same technical scoring as the sector orbs, one level down: to GICS sub-industry baskets built from S&P 500 members, and to DG individually.

Money Flow Index, 14 days (MFI) — a 0–100 oscillator computed over typical price × volume. Above 50 = net buying pressure; below 50 = net selling pressure.

Relative strength vs the S&P 500, 20 days — the stock's (or basket's) 20-session return minus SPY's over the same window.

Volume pressure — today's volume versus its own 20-day average, in standard deviations (z-score).

Industry baskets. Each industry's price series is the equal-weight average of its members' closes; volume is the members' summed volume. Formulas and thresholds are identical to the sector level.

Signals. Green = accumulation (MFI ≥ 60 and positive relative strength); red = distribution (MFI ≤ 40 and negative). A badge fires only when price also confirms: BUY = distribution + close below the lower 1σ Bollinger Band; CONTINUATION = accumulation + close above the upper 1σ band. No confirmation, no badge — same rule as the front page.

Win-rate cells. Each cell shows how often that signal has historically been followed by a rise over that horizon, computed on ~28 years of daily data (1998 → present) with no look-ahead — consecutive signal days count once, and the exit is the close of day t+h for each horizon (5–250 days). Cells also carry their sample size (n) and ±1 standard error (±SE). These are context, not proof — read them together with n.

The statistical bar. Two thresholds govern what "established" means here, both fixed in advance on ~28 years of daily data and requiring a win rate above 55% at 20 days. The sector-level publication gate (n > 500) is what a signal must clear before it's presented as proven on the front page — sector BUY (n = 2,403) and CONTINUATION (n = 3,501) pass it; read how the bar works. The sub-sector display gate (n > 300) covers industries and single stocks, which trade fewer signals than the 11 broad ETFs, so their buckets are smaller — at n = 300 the standard error of a 55–65% win rate is about ±2.9pp, still tight enough to be informative. Badges always show today's live signal regardless of sample size — the win-rate cells carry the evidence next to them. Buckets below the bar are research-grade context, not proven edges. Full tables in the back-test report.