Technical indicator analysis

The idea

Technical indicators are not magic — they are statistics. A relative-strength oscillator measures how one-sided recent closes have been. A moving-average convergence measure tracks the gap between fast and slow trend estimates. Volatility bands measure how stretched price is versus its own recent history. Each is genuinely informative some of the time, and famously misleading the rest of the time. The engineering problem is not computing them — it is knowing when each one deserves a vote.

Computed on your bars, not a lookalike feed

Indicators are computed on the exact bars rendered on your chart — the same symbol, timeframe and window — sourced from institutional-grade market data feeds, or, in screenshot mode, from the candles read off your image. Each indicator also declares its own warm-up requirement: a 50-bar average shown with 30 bars of data is a lie, so instead of showing one, the panel tells you it is still warming up.

The regime filter

The single most common way indicator systems lose money is using a mean-reversion tool in a trend, or a trend tool in a range. The engine measures trend strength directly (a directional-movement statistic) before weighting anything. In a strong trend, oscillators screaming “overbought” are down-weighted — overbought can stay overbought for weeks — and trend-following evidence is promoted. In a choppy range, the opposite. The regime classification itself is shown in the analyst report, so you can see which lens the engine applied and disagree with it.

Per-asset personality

Thresholds that are meaningful on an equity are noise on a memecoin. Volatility expectations, band widths and overbought/oversold levels are parameterized per asset class and per timeframe, which is also why the product insists you confirm what an uploaded screenshot actually shows before it analyses one.

One vote of four

The indicator channel produces one bounded score with its evidence attached. It is combined with pattern matching and sentiment under weights that are audited against real outcomes — no single indicator, however photogenic, can move a forecast on its own.