A technical indicator is a mathematical transform of price and/or volume data that summarizes one dimension of recent market behavior — trend, momentum, volatility, or volume participation — into a value that can be plotted, compared to thresholds, and read at a glance. Indicators exist because raw price is noisy: a transform (a moving average, a normalized oscillator, a range measure) imposes structure the eye and a rule engine can act on consistently. Their core tension is also their original sin: nearly all are derived from the same OHLCV series, so they tend to be lagging (they confirm what price already did) and mutually correlated (they re-describe one another). The discipline is therefore less about finding a magic transform than about choosing a small, complementary set and reading it honestly. This node defines the four families plus the cross-cutting confluence/divergence practice, and routes to each.

The four families

• Trend indicators — quantify direction and trend strength: MACD, ADX/DMI, moving-average systems (SMA/EMA crossovers, ribbons). Mostly lagging by construction (a moving average is a window over past prices). → Trend Indicators (MACD, ADX, MA Systems).
• Momentum oscillators — normalize rate-of-change into a bounded range (often 0–100) to flag overbought/oversold and divergence: RSI, Stochastics, CCI. Marketed as "leading," but they read off the same price and degrade badly in trends. → Momentum Oscillators (RSI, Stochastics, CCI).
• Volatility indicators — measure dispersion / range, not direction: Bollinger Bands (standard-deviation envelopes around an MA) and ATR (average of the true range). These don't say which way — they say how much movement, which is what risk framing needs. → Volatility Indicators (Bollinger Bands, ATR).
• Volume indicators — fold trade volume into price to gauge participation and conviction: OBV (cumulative signed volume), VWAP (volume-weighted average price), MFI ("volume-weighted RSI"). The one family whose input (volume) is not just a re-transform of price. → Volume Indicators (OBV, VWAP, MFI).

Confluence & divergence — the cross-cutting practice

Two universal techniques span all families. Confluence stacks multiple indicators (or indicators + structure) and acts only when they agree. Divergence flags when an indicator and price disagree — e.g. price makes a higher high while the oscillator makes a lower high — read as fading momentum. Both are covered in Indicator Confluence & Divergence. The critical caveat (below) is that confluence among same-source indicators is often not independent confirmation at all.

The honest meta-take (read this before stacking indicators)

Two facts should govern how the Augustus agent and any trader use this whole category:

1. Most indicators are lagging, correlated derivatives of one price series — so stacking them is redundancy, not confirmation. Trend-following indicators (moving averages, MACD) are defined as functions of past price and lag by construction; StockCharts' own ChartSchool and standard pedagogy classify them as lagging, with oscillators only loosely "leading." When three of your four indicators are computed from the same closes, their agreement is largely mechanical — it tells you the price did one thing, restated three ways. This is redundancy masquerading as confluence, and it manufactures false confidence. Genuine confluence requires orthogonal evidence (e.g. price structure + volume + volatility regime), not three momentum readings.

2. The standalone edge of indicator rules is weak and was largely arbitraged / data-snooped away. The landmark optimistic study — Brock, Lakonishok & LeBaron (1992) — found simple moving-average and trading-range-break rules on 90 years of Dow data produced returns inconsistent with random-walk, AR(1) and GARCH nulls, which looked like real predictability. But Sullivan, Timmermann & White (1999) re-examined that exact universe with White's Reality Check bootstrap, which adjusts for data-snooping (the fact that the "best" rule found by searching thousands of variants will look good by luck). They found the in-sample performance no longer significant after that correction, and reported low profitability in a subsequent ~10-year out-of-sample window, which they read as evidence of increasing market efficiency. The lesson generalizes: an indicator rule that backtests well is the survivor of an unstated search, and its edge tends not to persist out of sample.

The operational conclusion (not a claim that indicators are useless): they earn their keep as discipline, context, and risk-framing tools, not as alpha generators — enforcing consistent rules, defining a regime/context filter, and sizing and stopping trades. ATR is the most operationally valuable indicator in the set: Wilder designed it purely as a volatility measure, and its dominant modern use is non-directional risk control — ATR-multiple stops (commonly ~2×ATR) that scale to current noise, and volatility-based position sizing (smaller size when ATR is high). That role doesn't require ATR to predict anything, which is exactly why it survives the critique above. Whether a directional signal "works" on a given trade is a downstream judgment for Augustus (live data + Cairn's record), not something the indicator asserts on its own.

Strengths & limitations of the category

• Strengths: consistency (a rule engine reads an indicator the same way every time, removing discretionary drift), compression (one number summarizes a behavior), and — for volatility indicators specifically — quantified risk inputs that have real, non-predictive value.
• Limitations: lag and correlation (above); regime dependence — oscillators flag false "overbought" all the way up a strong trend, while trend indicators whipsaw in ranges; and the single most common misuse: treating multiple correlated indicators as independent votes and over-sizing on apparent "confluence."

System relevance

The Augustus trade-setup agent consumes indicator readings as context and risk inputs, not as standalone buy/sell signals. The hard caveat passed downstream: weight ATR-driven sizing/stops over directional oscillator signals, and never treat agreement among same-source indicators as independent confirmation. The four child nodes hold the per-indicator mechanics and defaults; this node and Indicator Confluence & Divergence hold the cross-cutting reading discipline.

Sources

• Brock, Lakonishok & LeBaron (1992), "Simple Technical Trading Rules and the Stochastic Properties of Stock Returns," Journal of Finance — original MA / trading-range-break result. https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1540-6261.1992.tb04681.x
• Sullivan, Timmermann & White (1999), "Data-Snooping, Technical Trading Rule Performance, and the Bootstrap," Journal of Finance 54(5):1647–1691 — Reality Check re-examination of BLL. https://onlinelibrary.wiley.com/doi/abs/10.1111/0022-1082.00163
• StockCharts ChartSchool, "Introduction to Technical Indicators and Oscillators" — leading vs lagging classification. https://chartschool.stockcharts.com/table-of-contents/technical-indicators-and-overlays/introduction-to-technical-indicators-and-oscillators
• Investopedia / OANDA on ATR (Wilder 1978) — volatility measure, ATR-multiple stops and volatility-based position sizing.