Levels, patterns and limits

This lesson brings together Fibonacci retracements, pivot points, divergence, candle shapes and multi-bar patterns. These tools make chart observations easier to describe. They do not create physical walls at particular prices. Buyers and sellers can quote and transact beyond a drawn line, and transaction costs can matter more than a visually neat pattern.

A repeatable description names the timeframe, data source, completed bars, boundaries and failure rule before the outcome is visible. A label added after a large move is not evidence that the same label could have predicted it. Keep observation, hypothesis and result in separate notebook columns.

The sequence 0, 1, 1, 2, 3, 5, 8, 13 and onward adds the previous two numbers to get the next. It is associated with Leonardo of Pisa, often called Fibonacci, whose 1202 Liber Abaci helped transmit the sequence and practical arithmetic in Europe. Related sequence work existed earlier in Indian mathematics. It is therefore misleading to say Leonardo invented every idea now associated with it.

Ratios between increasingly large neighbouring numbers approach about 0.618 in one direction and 1.618 in the other. Related ratios supply familiar 38.2% and 23.6% chart marks; 78.6% is commonly associated with the square root of 0.618. A 50% retracement is included by charting convention, not because one half is a Fibonacci ratio.

Borrowing a mathematical ratio does not prove an economic law. Traders can watch the same areas, but popularity alone does not establish dependable predictive power. Avoid claims that every currency swing must respect these percentages. Their useful beginner role is to make the distance from a documented swing explicit.

For an upswing, choose the low and high using a rule that existed before the later decline. Let the move be high minus low. A retracement down from the high at fraction r is high minus r times that move. For a downswing, a bounce measured upward from the low is low plus r times the decline’s distance. Reversing the anchors changes what 0% and 100% label on some platforms.

A platform may round to its quote precision. Rounding the arithmetic does not convert a line into a guaranteed execution price. A second learner choosing an earlier swing low will obtain other reference levels. Record dates and prices for both anchors, and do not keep moving them until a losing example looks successful.

A retracement measures a partial return through a prior move. An extension projects a distance beyond a selected endpoint, sometimes using a third anchor. Because platforms offer different tools, a label such as 161.8% is incomplete without its anchor convention. State the equation instead of relying only on a drawing.

Using a line as a hypothetical target also requires an entry and a failure level before a reward-to-risk ratio can be calculated. If no entry or invalidation is defined, calling an extension “excellent reward” is incomplete. Costs and available quote precision belong in the same plan.

A classic pivot uses the previous completed period’s high, low and close. The central pivot P is their sum divided by three. Classic first reference levels are R1 = 2P − low and S1 = 2P − high; R2 = P + (high − low) and S2 = P − (high − low). These are one family of formulas. Other pivot types can use different inputs or weights.

A day ending at midnight on one provider can differ from another provider’s daily boundary. Their highs, lows and closes can differ and so can their pivots. Using today’s unfinished high inside a formula for yesterday’s pivot is a different calculation. Agree on the session, timezone and pivot family before comparing numbers.

A common bearish divergence definition is a higher price swing high accompanied by a lower oscillator high at the corresponding observations. A common bullish divergence compares a lower price low with a higher oscillator low. These descriptions concern disagreement between selected swings; the word divergence does not guarantee a trend reversal.

Some traders discuss hidden divergence with different high/low relations. Definitions vary, so label the version instead of assuming every use of the word means the same pattern. Comparing a price peak from Tuesday with an unrelated oscillator reading from Friday manufactures an apparent disagreement. Match the observations deliberately.

A candle records open, high, low and close for a chosen interval. The real body spans open to close; wicks show the remaining range. A green candle commonly has close above open and red close below open, but platform colours are configurable. The shape does not show the exact order of all intrabar moves or the number of participants.

Modern candle analysis is associated with historical Japanese market practices. Munehisa Homma is often linked to early rice-market analysis, but attributing the exact modern OHLC candle to one documented invention is too confident. English-language popularisation came much later. Knowing this history does not increase a pattern’s reliability by itself.

Before naming a candle, record its body size relative to its whole range, the lengths of both wicks and the prior price context. A hammer-shaped candle after a decline is conventionally discussed differently from the same shape after a rise. “Small” and “long” need a numerical rule if you want another learner to reproduce the classification.

A doji is not proof that every buyer and seller has become equally uncertain; it records nearly matching open and close. A marubozu does not prove price moved only one way during the interval: four OHLC values cannot reconstruct the entire path. Describe what the data records before adding a psychological story.

Engulfing usually refers to the real bodies, not necessarily the entire wick-to-wick range. Define whether equality at a boundary counts. Harami describes a smaller nested body; direction, gap requirements and thresholds can vary between references. A name alone is not a reproducible trading rule.

Near-continuous Forex quotes often show textbook opening gaps less neatly than session-based stock charts. If you relax a gap condition for Forex, write the adapted definition and test that rule. Do not claim a strict textbook pattern appeared when one of its required conditions was absent.

These bring the practical candle vocabulary to sixteen core examples. It is not a list of every name ever used. “Strong,” “pause” and “advancing” need thresholds and close/open relationships for a repeatable test. Some definitions also require gaps, wick restrictions or a close into the first body. Record the version rather than merging incompatible descriptions.

A third candle completes a three-candle pattern only after its needed data becomes available. You cannot count its eventual close as information known at the second candle. Waiting for another observation may reduce one uncertainty while giving a later entry; confirmation is a trade-off, not free certainty.

A potential double top is not a completed reversal just because two highs look similar. A rule might require a completed close below the intervening low, with tolerance for peak separation and price difference. Another rule might use an intrabar breach. Each produces a different sample. A failed confirmed pattern also needs a definition, such as return above a stated boundary.

Pattern targets based on height are projections under a convention, not promises of movement. Changing timeframe can hide or create a shape because bars aggregate different intervals. A line drawn through chosen extremes is subjective unless the selection rule is explicit.

Several tools agreeing may be correlated rather than independent. Fibonacci, pivots, a moving average and a pattern neckline can cluster near one price, but the cluster is not proof of a barrier. Ask which inputs each tool used and what genuinely new information was added. Then record a plausible alternative outcome.

Separate a plan’s arithmetic from its evidence. Test the unchanged rule on later examples, retain failures, avoid counting overlapping trades as independent events and account for costs. Optimising many settings until old data looks good is overfitting. A short successful sample is not a guarantee for future prices.

Finally describe a pattern without using its name: specify prices, dates, candle relations, tolerance, prior trend, timeframe and completion rule. Ask whether another learner can reproduce it. If not, improve the definition before assessing results. Level 5 will build on these limits by making the planned cash loss and position size explicit.