Technical analysis (TA) is the study of historical price and volume — almost always via charts — to inform forecasts and trading decisions. It rests on three working premises: that price discounts known information, that trends persist long enough to trade, and that history rhymes because human crowd psychology repeats. None of these is self-evidently true; all three sit in direct tension with the Efficient Market Hypothesis (EMH), which holds that past prices contain no exploitable forecasting information. That tension — between a vast practitioner tradition and a skeptical academic literature — is the defining feature of the field, and this overview's job is to map the branches honestly and deliver a balanced meta-verdict, not to defend or dismiss TA wholesale.

The major branches (and where to route)

The discipline fans out into roughly fifteen families, from foundational theory to esoteric forecasting systems:

• Dow Theory & Foundational Principles — the 19th-century origin (trend stages, confirmation, volume as confirmation) that frames everything below.
• Chart Types — line, bar, candlestick, point-and-figure, Renko, Heikin-Ashi: the visual substrate.
• Trend Analysis — trendlines, channels, moving averages, the "trend is your friend" core.
• Support & Resistance — price memory, prior highs/lows, round numbers, supply/demand zones.
• Chart Patterns — head-and-shoulders, triangles, flags, double tops (see Bulkowski for measured base rates).
• Candlestick Patterns — engulfings, dojis, hammers; Japanese single/multi-bar formations.
• Technical Indicators — RSI, MACD, stochastics, Bollinger Bands and the broader oscillator/overlay toolkit.
• Volume & Order Flow — volume confirmation, OBV, VWAP, tape/footprint reading.
• Elliott Wave — five-up/three-down fractal wave counting.
• Fibonacci — retracement/extension ratios derived from the golden-ratio sequence.
• Harmonic Patterns — Gartley, Bat, Butterfly; Fibonacci-defined geometric setups.
• Gann Theory — angles, squaring price and time, cycles.
• Relative Strength & Rotation — cross-asset/cross-sector leadership ranking (distinct from the RSI indicator).
• Multi-Timeframe Analysis — aligning signals across horizons.
• Market Breadth & Internals — advance/decline, new highs/lows, % above MA; the "health" of an index beneath its price.

Each branch has its own section node; consult those for mechanics, defaults, and measured base rates. This doc deliberately stays at altitude.

The honest meta-verdict

This is the heart of the doc, so it must not be smoothed over. The rigorous, out-of-sample, cost-and-data-snooping-adjusted evidence is weak-to-null for the large majority of TA — most chart patterns, most candlestick patterns, most oscillators, and the esoteric systems (Elliott, Gann, harmonics) — once you account for two things that destroy most "edges": data-snooping (test enough rules and some will look profitable by chance) and real-world trading costs. The canonical critique is Sullivan, Timmermann & White (1999): using White's bootstrap "Reality Check" across a large universe of rules on ~100 years of Dow data, the best-performing rule was statistically superior in-sample — but it did not survive out-of-sample, and broader work in that line finds no robust profitable simple rule on the major US indices once snooping is corrected. The reproducibility problem is structural: the field generates rules far faster than markets can validate them, so most "what works" claims are survivorship and overfitting laundered into folklore.

But intellectual honesty cuts both ways — there are genuine, academically-supported exceptions that deserve real weight, and it would be folklore in the other direction to deny them. Two stand out:

1. Momentum / relative strength. Jegadeesh & Titman (1993) documented robust cross-sectional momentum: buying recent winners and shorting recent losers (3–12 month formation) earned roughly 1% per month on average, not explained by systematic risk. Moskowitz, Ooi & Pedersen (2012) documented time-series momentum across ~58 futures markets in four asset classes — persistent trends over 1–12 months that partially reverse later. These are among the most replicated anomalies in finance.

2. The price–volume relationship. Karpoff's (1987) survey established that volume is positively related to the absolute magnitude of price change (and, in equities, to price change per se) — a robust, well-documented contemporaneous regularity.

So the defensible reading of TA is twofold. First, a small handful of real edges (momentum/RS, and volume as an information/confirmation proxy) that the academic literature actually supports — distinct from the toolkit's folklore. Second, and arguably more durable, TA as a risk-management and trade-structuring framework: defining entries, explicit invalidation levels, ATR-based stops and position sizing, and regime/breadth context. That value holds regardless of forecasting power — a well-placed stop is useful even if the entry signal is coin-flip. What TA is not is a crystal ball, and any branch that markets itself as one (precise Elliott counts, Gann time-squaring, harmonic price targets to the cent) should be treated as low-confidence at best.

Standing & contestation

TA's standing is genuinely split by audience. It is near-universal among retail and discretionary traders and embedded in execution and risk tooling institutionally; the forecasting claims, meanwhile, are largely rejected by mainstream academic finance. Both positions contain truth, and the corpus should refuse to flatten either — be fair to skilled practitioners (whose edge often lives in discipline and risk control more than in any pattern), while refusing to present unproven signals as established fact.

System relevance

For the Delvantic pipeline and the Augustus trade-setup agent, this verdict translates into an explicit confidence hierarchy. Weight the academically-supported layers: momentum / relative strength, the volume–price relationship, breadth/regime context, and the risk-framing apparatus (invalidation, ATR stops, sizing). Treat pattern, candlestick, oscillator, and esoteric (Elliott/Gann/harmonic) signals as low-confidence confluence at most — they may add color but should never carry a thesis alone. Cairn's measured track record, not this doc, is the final arbiter of what is actually working in live conditions; this overview simply sets the prior. The branch nodes below supply the mechanics for each tool; consume them through this lens.

Sources

• Sullivan, R., Timmermann, A., & White, H. (1999). Data-Snooping, Technical Trading Rule Performance, and the Bootstrap. Journal of Finance, 54(5), 1647–1691. — https://onlinelibrary.wiley.com/doi/abs/10.1111/0022-1082.00163
• Jegadeesh, N., & Titman, S. (1993). Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency. Journal of Finance, 48(1), 65–91. — https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1540-6261.1993.tb04702.x
• Moskowitz, T. J., Ooi, Y. H., & Pedersen, L. H. (2012). Time Series Momentum. Journal of Financial Economics, 104(2), 228–250. — https://www.sciencedirect.com/science/article/abs/pii/S0304405X11002613
• Karpoff, J. M. (1987). The Relation Between Price Changes and Trading Volume: A Survey. Journal of Financial and Quantitative Analysis, 22(1), 109–126. — https://ideas.repec.org/a/cup/jfinqa/v22y1987i01p109-126_01.html
• Fama, E. F. (1970). Efficient Capital Markets: A Review of Theory and Empirical Work. Journal of Finance, 25(2), 383–417. (EMH framing.)