Consider a hypothetical GBP/USD chart with repeated contacts near a predefined resistance zone. A trader plans a short entry and a stop, then checks whether the observed prices actually support the rule. A chart pattern describes a sample of price history; it does not establish the next move or a profitable outcome. This article introduces the assumptions and tools of technical analysis.

What technical analysis really is

Technical analysis organizes observations about price, changes over time and available market activity. A chart can describe a trend, mark a zone and define a condition to test. Its appearance alone does not supply a reliable probability for the next move. Profitability belongs to specified trading rules, data and costs; a price-based observation does not establish that every economic fact is already perfectly priced.

One historical framework is Dow theory, developed from Charles Dow’s observations and later work by William Hamilton and Robert Rhea. StockCharts explains its original comparison of the industrial and rail equity averages. That history does not make every modern chart rule a Dow theorem or show that an FX rule works. Treat the framework’s assumptions separately from evidence about a particular strategy.

Popular levels may influence some participants’ decisions, but their popularity does not reveal how many will buy or sell, available execution prices or the next direction. Market turnover and liquidity are not evidence that a technical method has an edge. A forecast still needs an operational definition and a test, including cases where the proposed level did not hold.

Three common premises to examine

Technical-analysis introductions often discuss price incorporating information, trends and recurring patterns. These are ways to organize observations, not results of a test. Dow theory’s historical assumptions and theorems have a more specific equity-market context.

Price and information. The proposition that price reflects available information does not prove instant, complete or error-free pricing. Nor does it establish that fundamental information is useless. For a test, distinguish the time a price or announcement became available from the time it was later stored or revised.

Trends and horizon. Historical Dow theory distinguishes primary, secondary and daily movements; its duration descriptions are guidelines, not universal forecasting limits. On your chosen horizon, specify how highs and lows are selected and when they become confirmed. A shorter-term rise can coexist with a longer-term fall. Record the M1, H1 or D1 timeframe instead of assigning one compulsory horizon to each trading style.

Recurring shapes. Similar-looking candles or levels may occur again, but appearance does not imply the same outcome. A psychological explanation for a pattern is a hypothesis, not proof that participant behaviour is invariant. Define the pattern before selecting examples and retain failed as well as successful cases.

Four tools and the questions they answer

Four commonly used tools provide a starting vocabulary. Their presence on a chart does not establish a profitable edge; each complete trading rule needs testing with costs, sample definitions and out-of-sample data.

Trend lines connect selected price points. Two points determine a straight line; further contacts can be observations under a stated selection rule, rather than a guarantee of continuation. Specify which highs or lows count and when they were known. Do not redraw the line after every failed contact merely to fit the history.

Support and resistance are proposed zones of reaction, often based on selected highs, lows or other price references. CME stresses that they do not have to hold to the smallest price increment. A former support can be investigated as resistance after a break, but a role change is not guaranteed. Record your own level-selection rule instead of assuming that every participant uses the same prices. To describe successive confirmed points, see how to read market structure.

OHLC candles record open, high, low and close for a selected period. The body joins open and close; the wicks extend to high and low. A bullish candle means close exceeds open, not that its next candle must rise. As a hypothetical EUR/USD calculation with one pip = 0.0001, O 1.1000, H 1.1030, L 1.0990 and C 1.1020 give a 40-pip range, a 20-pip body and two 10-pip wicks. Check the data source: MT5 documents Bid-based OTC charts, while purchases use Ask; received tick counts are not global FX turnover.

Japanese candlestick anatomy — bullish candle Anatomy of a single Japanese candlestick: body defined by open and close, upper wick from highest price to top of body, lower wick from lowest price to bottom of body. high low open close body upper wick lower wick forex-podstawy.pl
Figure 1. Anatomy of a bullish Japanese candlestick: body defined by an open below and a close above, upper wick from the period high to the top of the body, lower wick from the low to the bottom.

Moving averages transform selected historical observations. SMA assigns equal weights to its window; EMA gives more weight to recent inputs. Hypothetical closes 1.1000, 1.1010 and 1.1020 give SMA(3) = 1.1010. After a new close of 1.0990, the oldest value drops out: (1.1010 + 1.1020 + 1.0990) / 3 ≈ 1.10067. Price below the average or a crossing of two averages is a defined condition, not proof of a profitable short or a guaranteed change in trend.

Practical takeaway: Technical analysis uses selected historical prices and sometimes volume. A pattern description alone does not establish predictive accuracy.

OHLC does not order the level touches

Four prices can describe different paths within the same candle. See why testing orders inside one bar requires an explicit sequence rule.

Scroll sideways to see the full diagram

OHLC does not order the level touchesTwo hypothetical paths have the same O 100, H 104, L 96, C 101. A long opens at 100 with TP 103 and SL 98. Visiting the high first closes at TP for +3; visiting the low first closes at SL for −2. Later prices cannot change an already closed result.One OHLC candle, two possible sequencesA · High firstTP 103O 100SL 98OpenCloseB · Low firstTP 103O 100SL 98OpenCloseA: close +3 · B: close −2Both: O 100 · H 104 · L 96 · C 101

Hypothetical model: one instrument unit, continuous paths between points, ideal execution at the first level, no spread or costs. Results do not describe real order execution. TradingView explains that historical OHLC does not retain the full intrabar path.

The same candle may not identify which order filled first.

When was the signal actually available?

On an open bar, high, low and current price can change. An indicator may cross a threshold and return before the close. TradingView calls differences between historical and realtime calculations or plots repainting; not every update is an error. State whether a rule uses intrabar values or waits for a confirmed close.

A line depends on the chart scale

Hypothetical prices 100, 200 and 400 occur at equal time intervals. Connect the first and last points. On a linear scale, the line’s midpoint is (100 + 400) / 2 = 250, so the observed 200 lies below it. On a logarithmic scale, the midpoint corresponds to √(100 × 400) = 200, so the same observation lies on the line. Both successive changes are +100%. The data and anchor points are identical; only the vertical transform changes. Record the scale before claiming that price broke a drawn line. Neither scale guarantees a better forecast.

A pivot can be plotted on an earlier bar even though recognizing it required later observations. Use it in a test only from the time the necessary data became available. Likewise, the final daily high cannot be known that morning. Such future-data leakage can make historical results appear much better than a method that could actually have been executed.

A trading test needs entry, exit, sizing and execution rules as well as a chart condition. Check the instrument and pair characteristics and translate the planned distance using the pip value at the selected position size. Include commission, financing where applicable and slippage. Actual Ask-entry and Bid-exit prices already contain the spread; do not deduct it twice. Ordinary stop execution can differ from its planned level.

What this means for a beginner trader

Knowing a candle’s anatomy or a moving-average formula is a useful skill, but does not rank you above other beginners or prove a complete trading method. Start with one observation you can define and reproduce. Add a tool when its calculation answers a specified question, rather than follow a compulsory sequence of indicators.

An RSI threshold or an indicator crossing can be part of an explicit rule; by itself it does not establish profitable execution. Specify the instrument, data frequency, confirmation time and complete trade rules. Evaluate the result after costs rather than label every isolated indicator either a guaranteed signal or mere noise.

A concise reference on chart construction and the connection between technical setups and broader market conditions is available in the Technical Analysis section on ForexMechanics, with deeper coverage of trend mechanics in the trend glossary entry.

What to do tomorrow

  1. Describe one closed candle. Record its instrument, source, timestamp and O/H/L/C. Compute body and range without predicting the next candle.
  2. Define a zone or line. Write down which confirmed points qualify, the chart scale and the rule for a break. Keep examples that do not react as expected.
  3. Reproduce one calculation. Compute a short SMA by hand, then compare the platform’s input price and window. Record any initialization or data difference.
  4. Check signal availability. Distinguish current-bar values from confirmed closes and pivots requiring later bars. OHLC alone does not order intrabar touches of both SL and TP.
  5. Separate learning from testing. A test needs complete entry, exit, size, cost and fill assumptions. Fix rules on one sample and evaluate a later unused sample without retuning; retain poor results. Demo work can reveal problems but cannot guarantee future profit. You can continue learning without risking money.