CompanyGraph
What price, volume, and volatility can show about market behaviour—and what they cannot predict.
A Chart Records Behaviour, Not Destiny
Technical analysis starts with traded prices, transaction volume, and time. A moving average summarizes past prices. A volatility measure summarizes their dispersion. A support level names a zone where trades previously concentrated. These are observations produced from the market record.
The error is not drawing a line; it is silently turning the line into a cause or a forecast. A share price above its 200-day average establishes that relationship for the selected data and date. It does not establish that the company is undervalued, that buyers will continue to appear, or that the trend will persist.
What the Main Tools Measure
Trend measures such as moving averages and higher-high/higher-low rules reduce short-term noise and classify direction over a defined window. The choice of window changes the classification. A 20-day trend and a 200-day trend can both be correct observations of different horizons.
Volume measures record executed trading under a market’s reporting rules. Rising volume with a price advance may indicate broader or more urgent trading participation; it does not reveal whether the trades were informed, hedging, forced, or simply reallocations.
Volatility measures describe the size and frequency of observed price changes. Compression can precede a larger move, but the direction and timing remain unknown. Options-implied volatility adds a market price for insurance, not a guaranteed forecast.
Momentum and divergence compare recent returns or an indicator with price. They can flag that price and one derived measure are moving differently. That mismatch is a condition to investigate, not proof that a reversal must follow.
Provenance and Evidence
Technical practice descends from charting traditions and Dow Theory, while modern research tests particular rules rather than validating a single universal method. Lo, Mamaysky, and Wang examined whether computationally defined technical patterns contained information about returns; their paper is evidence about a method and sample, not a license to treat every visual pattern as predictive. Their 2000 study makes the definition of a pattern part of the result.
Earlier studies of moving-average and trading-range rules reported return differences in historical U.S. data, but those results are sensitive to sample period, transaction costs, shorting, liquidity, and data-snooping. Brock, Lakonishok, and LeBaron provide a documented test, not a universal edge.
Why Apparent Signals Change
Once a rule becomes widely used, its trading can alter the very price behaviour that created the historical pattern. Market structure also changes through electronic execution, fees, tick sizes, short-sale rules, and the composition of participants. A rule that works in a liquid index may fail in a thin stock because entering and exiting moves the price.
Backtests can overstate a signal through look-ahead bias, survivorship bias, parameter selection, unrecorded delistings, and execution assumptions. A strategy that is profitable before costs may be unusable after spread, market impact, borrowing expense, tax, and delay. The investor should inspect the exact universe, dates, signal definition, and trade assumptions.
Technical Observations Meet Business Reality
Price and volume are not independent of the company. A product failure, refinancing need, regulatory action, or change in customer demand can alter the flow of orders. The chart may show the market’s response before the annual report explains it, but the chart does not identify the underlying event.
Conversely, a strong chart can reflect a temporary liquidity wave while the business deteriorates. Fundamental analysis can test margins, cash conversion, debt, and operating capacity; technical analysis can describe how market participants are currently pricing those facts. Combining them is a choice of evidence, not proof that one is the cause of the other.
How to Use It Responsibly
- State the horizon and the exact rule before looking at the outcome.
- Test the rule on a population that includes delisted and failed names where possible.
- Separate signal frequency from payoff size and include realistic trading costs.
- Ask what market mechanism could make the pattern persist and what would make it disappear.
- Use a chart to trigger investigation or define a decision rule, not to claim certainty about the future.
Technical analysis is most defensible as a disciplined language for observed market behaviour. Its value depends on transparent definitions, honest testing, and a clear boundary between a recorded pattern and an inferred forecast.