Technical Analysis: Charts, Indicators and Their Limits

On 20 July 1901 the Wall Street Journal ran a Charles Dow editorial called “Methods of Reading the Market,” which set out the “book method” of recording prices in columns, the ancestor of today’s point and figure chart. Six months later he went further. In January 1902 Dow wrote that a bull period lasts while each high in the average clears the highs before it, and a bear period starts when lows begin undercutting earlier lows. A CFA Institute Research Foundation review from November 2016 cites both editorials. The 1902 rule still defines an uptrend. Almost every chart tool invented since is a faster way to read that same record.

What does a technical analyst assume?

The method rests on a few assumptions. Each fails under a known condition.

Price is assumed to reflect what is known. Earnings expectations, interest rates, rumors and fear all turn into buy and sell orders, so the chart is treated as a running summary of everything the crowd has acted on, right up to the last trade. That holds until news arrives that nobody had. A guidance cut announced overnight is in no chart until trading resumes.

Prices are also assumed to trend. A stock that has been rising is treated as more likely to keep rising than to reverse at random, which is why most chart rules follow the trend and why they bleed money, a little at a time, when the stock settles into a range.

Behavior repeats, too. Fear and greed draw similar shapes year after year. So a pattern that came before a move once may come before one again, although the shapes are drawn by eye and two analysts can label the same bars differently.

Which chart types do analysts read?

Chart	What each bar shows	Good for
Line	One price per period, usually the close	Seeing the trend without intraday noise
Bar (OHLC)	Open, high, low and close on one vertical line	The range and where price settled
Candlestick	The same data, with a shaded body between open and close	Reading who won the session
Point and figure	Columns of X’s and O’s that move only when price moves a set amount	Clean levels and breakouts, with time ignored

Bars and candles carry identical data. Use whichever you read faster. Point and figure, the descendant of Dow’s book method, earns its place for one job: it strips time out of the chart, so a level that price has hit again and again stands out without a crowd of small bars around it.

How do you tell the trend?

Start with swing points. An uptrend is a run of higher highs and higher lows, which is Dow’s 1902 rule applied to a single chart, and a downtrend is the mirror image. A lower low after a series of higher lows puts the uptrend in question. A lower high usually shows up first, as a warning. Dow’s successor at the Journal, William Peter Hamilton, kept writing in the same vein, and Robert Rhea later gathered both men’s editorials into a book called The Dow Theory.

Moving averages describe the same thing with less judgment. A rising 50-day average with price above it is an uptrend in one line on the chart, and when price keeps crossing back and forth over a flat average, the stock is in a range, where trend rules get whipsawed and range tactics, buying near support and selling near resistance, take over.

Where do support and resistance come from?

Support is a price where declines have stopped before. Resistance is where rallies have stalled. They form at old swing points and gap edges. Round numbers and long moving averages attract them too.

Treat them as zones a little wider than a single price. A broken level often swaps roles: resistance that gives way on a strong close tends to act as support on the next pullback, because traders who sold too early and buyers who missed the move both have orders waiting there.

Investors who already own shares use the same levels. Someone writing a covered call against a holding often sets the strike just above resistance, on the reasoning that a level that has capped rallies for months may cap the next one before expiration.

What do the indicator families measure?

Family	Examples	Question it answers
Trend	Simple and exponential moving averages, MACD	Which way is price moving, and how steadily?
Momentum	RSI, stochastic oscillator, rate of change	Is the move gaining or losing push?
Volatility	ATR, Bollinger Bands	How far does price normally travel in a bar?
Volume	Volume bars, on-balance volume, VWAP	How much participation sits behind the move?

Every indicator is a formula applied to price or volume. The relative strength index, developed by J. Welles Wilder, is RSI = 100 - 100 / (1 + RS). RS compares up-closes with down-closes: mean gain over mean loss across the chosen window. Pick one tool per family. Two momentum oscillators on the same chart mostly agree, since they are built from the same closes, and the best swing trading indicators are the ones that each answer a different question.

How do chart patterns and volume fit together?

Patterns sort into continuation shapes, such as flags, pennants and the cup and handle, and reversal shapes, such as head and shoulders and double tops and bottoms. A pattern is only a sketch until price closes beyond its boundary. That close is the breakout.

Volume is the check. A breakout on volume well above its recent average shows that many traders acted at that price, while a breakout on thin volume leaves the move resting on a handful of orders and fails more easily when the first sellers show up. Falling volume during a pullback inside an uptrend reads as a pause. Rising volume on the way down reads as distribution.

Which timeframe fits your holding period?

Match the chart to how long you hold. Day traders work from minute charts. Swing traders hold for days to weeks and read the daily. Investors use weekly and monthly bars.

Shorter charts produce more signals and more false ones. Random noise makes up a larger share of a five-minute bar than of a weekly one, which is why many traders read top-down: the weekly sets the direction, the daily picks the entry, and if the two disagree they size down or wait.

Which price history is your chart drawing?

Most chart data is adjusted for stock splits. Many feeds also adjust for dividends. Both adjustments rewrite old prices, so a level drawn on bars from two years ago may sit at a price that never actually traded, and two data sources can show the same support at different numbers.

On a price-only chart that old ceiling still reads $50. Orders fill at real prices. Set alerts and stops against the series you actually trade on. Funds that keep losing value face the same issue harder: many of the products on an inverse ETF list reset daily, and a fund that reverse splits rescales all of its history at once.

How does a chart turn into an order?

Take a hypothetical stock that ran from $50 to $60. It trades above a rising 50-day average, then pulls back toward the middle of the run, and a Fibonacci retracement of 38.2% marks one place where buyers who missed the first leg might step in.

The chart gave the stop. The risk budget turned it into a share count. The position size calculator runs the same sum for any account. That position is 56.5% of the account, because a tight stop lets the share count balloon, so concentration needs a separate cap set by how many holdings you want, which is the subject of how many stocks you should own. If price closes below $55.50, the setup is off.

Technical analysis or fundamental analysis?

Fundamental analysis asks what a company is worth. It works from revenue, margins, debt and cash flow. Technical analysis asks what the stock is doing and where the trade is wrong. Many investors use both. A common split is to let earnings, the balance sheet and valuation decide which companies deserve money at all, and then let the chart decide the day to buy, how many shares, and the price at which the position gets cut.

What can a chart not tell you?

A chart cannot value a business, and it cannot see a court ruling or a merger before the headline; when news like that lands overnight, the history on the screen stops describing the present.

Patterns are subjective. Indicators lag, because they are averages of the past. Trend signals misfire in sideways markets. And a level watched by enough traders can work partly because they all act on it, which says nothing about whether it will hold next time.