Shooting Star Candlestick: Rules and Confirmation

A shooting star candlestick is a candidate bearish reversal pattern considered after a defined price advance. It normally has a small body near the lower part of its range, a long upper wick, and little lower wick. The candle shows that price traded above the body and finished below the high. It does not prove why the move occurred or guarantee that price will decline next.

To use the pattern consistently, define five things before reviewing the outcome: the candle geometry, the prior advance, the chart and session settings, the confirmation event, and the failure or expiry rule. Without those definitions, almost any long-upper-wick candle can be labeled a successful shooting star after the full chart is visible.

Educational note: This guide explains pattern classification and historical testing. It is not a trade recommendation, a fixed entry/stop/target system, or evidence that a single candle predicts future returns.

Key Takeaways

A long upper wick is not automatically a shooting star; a written prior-advance rule is required.
Open, high, low, and close can be converted into exact body and wick measurements.
A still-forming candle is provisional and can lose the pattern before the interval closes.
Shooting star, inverted hammer, gravestone doji, hanging man, and evening star are different classifications.
Confirmation and invalidation are later events, not hidden inside the candidate candle.
No universal win rate, best timeframe, volume threshold, or confirmation rule applies to every market.
Replay practice is useful only when definitions are frozen before future candles are shown.

Which Candlestick Guide Should You Use?

ChartMini has several candlestick pages. Their roles should remain separate.

Question	Primary guide
How do open, high, low, close, bodies, wicks, sessions, and live candles work?	Complete candlestick chart reading workflow
What are the main named single- and multi-candle patterns?	Candlestick patterns guide
How should general pattern filters and confirmation variables be compared?	Candlestick setup and confirmation checklist
How do I define and test one shooting star candidate?	This guide
How does the three-candle evening star differ?	Evening Star guide
How do I identify a historical resistance candidate without hindsight?	Support and resistance identification checklist
How can I practice pattern recognition with the right edge hidden?	Pattern recognition with chart replay

This page therefore owns the exact Shooting Star definition, context, confirmation, failure, versioning, worksheet, and replay-review workflow. It does not own the full candlestick catalog or a universal trading strategy.

What a Shooting Star Actually Records

A standard candlestick contains four prices for one interval:

O = open
H = high
L = low
C = close

The candle's geometry can be calculated directly:

Range = H - L
Body = absolute value of (C - O)
Upper wick = H - max(O, C)
Lower wick = min(O, C) - L

A shooting-star-shaped candle generally has:

a body near the lower part of the total range;
an upper wick that is large relative to the body;
a small or absent lower wick;
a completed candle;
a predefined advance before the candle.

These are descriptive conditions. The OHLC data show where the interval opened, traded, and closed. They do not reveal the exact chronological path between the high and low, the identity of participants, unfilled orders, or what the next interval must do.

One reference definition for testing

Different books, platforms, and scanners use different thresholds. To make a replay study reproducible, you could begin with a reference definition such as:

Range > 0
Upper wick >= 2 × Body
Body share of range <= 35%
Body is located in the lower 40% of the range
Lower wick <= Body
Prior advance rule = true
Candle status = closed

This is a test specification, not a claim that these values are optimal. A stricter project might require a longer upper wick or a smaller lower wick. A broader project might allow more variation. Record the version before testing and do not change thresholds because one historical example looks inconvenient.

Handle near-zero bodies separately

If O and C are nearly equal, the body approaches zero. The ratio Upper wick ÷ Body can then become extremely large or undefined.

Use one of these controls:

set a minimum body size in ticks or percentage of range;
classify near-zero bodies as a separate gravestone-doji candidate;
use range-share thresholds instead of dividing by a near-zero body;
record both categories and compare them separately.

Without this control, a scanner can label almost any long-upper-wick doji as an infinitely strong shooting star.

Verify the Chart Before Measuring the Candle

Two charts can display different shooting-star candidates even when the symbol looks the same. Freeze the chart identity first.

Field	What to record	Why it matters
Instrument	Symbol, venue, contract, expiry, or provider product	Similar symbols can represent different markets
Price source	Last trade, bid, ask, midpoint, mark, index, or settlement	OHLC values can differ by source
Session	Regular, extended, overnight, continuous, or custom	Session boundaries change candle construction
Timezone	Exchange time, UTC, local, or provider time	Daily and weekly candles can open and close differently
Timeframe	Exact interval or bar type	One-minute and daily patterns summarize different data
Adjustment	Raw, split-adjusted, dividend-adjusted, or futures-roll method	Historical highs and bodies may change
Volume type	Exchange, consolidated, broker, tick, or unavailable	Volume comparisons are not interchangeable
Bar state	Closed or still forming	A live high, low, and close can continue changing

A pattern detected on a regular-session stock chart may disappear when extended-hours data are included. A daily Forex candle can change when the provider uses a different timezone. These are data-definition differences, not proof that one chart is wrong.

Define the Prior Advance Before Naming the Pattern

The same candle shape is called a shooting star after an advance and an inverted hammer after a decline. “It looks like price went up” is not a reproducible context rule.

Possible prior-advance definitions include:

Pivot-sequence rule

Require a written sequence such as higher confirmed swing highs and higher confirmed swing lows before the candidate candle.

Controls to record:

pivot lookback on each side;
equality tolerance;
candidate versus confirmed pivots;
local versus major structure;
last eligible pivot before the candle.

Net-change rule

Require the close before the candidate to be above the close a predefined number of bars earlier.

Prior change = Close[t-1] - Close[t-n]

The project must choose n, the minimum change, and whether the measure is absolute, percentage-based, or volatility-normalized.

Range-break rule

Require price to have closed above a previously defined range or swing boundary before the candidate appears.

Slope or average rule

Require a predefined positive slope or price position relative to a moving reference. If used, record the calculation and lookback. Do not call an indicator an independent confirmation if it is derived from the same price sample without acknowledging the overlap.

No method is universally best. The purpose is to prevent a long upper wick from being renamed after the later decline makes the preceding movement look obvious.

Closed Candle vs Live Candle

A shooting star should be classified using a closed candle unless the project explicitly studies live-bar signals.

During formation:

the high can rise;
the low can fall;
the current close can move;
the body can expand or shrink;
the upper-wick ratio can change;
a shooting-star shape can become a large bullish candle before the interval ends.

If live-bar detection is part of the study, record snapshots separately. Do not compare provisional live shapes with final closed-candle patterns as if they were the same event.

Shooting Star vs Similar-Looking Patterns

Pattern	Geometry	Required context	Main distinction
Shooting star	Small lower body, long upper wick, little lower wick	After a predefined advance	Single-candle bearish-reversal candidate
Inverted hammer	Often the same geometry	After a predefined decline	Single-candle bullish-reversal candidate
Gravestone doji	Open and close nearly equal near the low, long upper wick	Context varies by framework	Near-zero body requires separate tolerance
Hanging man	Small upper body with a long lower wick	After an advance	Lower-wick shape, not upper-wick shape
Evening star	Bullish candle, small middle candle, bearish third candle	After an advance	Three-candle sequence with its own completion rule
Random upper-wick candle	May resemble a shooting star	No qualifying prior advance	Shape alone does not meet the definition

Shooting Star vs Inverted Hammer

The candle color and wick geometry can be identical. The label changes because the prior-price context changes.

A robust dataset should not contain a single field called “upper-wick reversal.” Record at least:

direction of the qualifying prior move;
pattern label;
chart timeframe;
geometry version;
confirmation version;
outcome horizon.

Shooting Star vs Gravestone Doji

Define what counts as “open and close nearly equal.” Possible tolerances include:

a maximum number of ticks;
a maximum percentage of the total range;
a maximum percentage of price;
the platform's documented doji rule.

Do not decide whether the candle is a doji only after seeing which label produces a better historical result.

Shooting Star vs Evening Star

A shooting star is complete after one qualifying closed candle. An evening star normally requires three candles, with the third candle providing the bearish response. A shooting star can appear as the middle candle of an evening-star-like sequence, but the two events should be stored separately because their confirmation timing differs.

Location Is a Recorded Variable, Not Proof

A shooting star is often reviewed near resistance, a prior high, or a failed breakout. These locations may be useful study variables, but they require definitions.

Possible location categories include:

prior confirmed swing high;
upper edge of a predefined range;
previously drawn resistance line or zone;
prior breakout boundary;
session, week, or month high;
calculated reference such as a moving average or Fibonacci level;
no predefined reference nearby.

For raw-price references, use the support and resistance candidate-identification workflow. Keep calculated references in a separate field so a Fibonacci or moving-average value is not rewritten as historical resistance.

A long upper wick near a level proves only that the interval traded above its body and closed below its high. It does not prove:

that institutions sold the high;
that late buyers are trapped;
that market makers hunted stops;
that unfilled supply remains at the level;
that the next candle will be bearish.

Those stories require separate evidence. The OHLC candle alone cannot identify participants or motives.

Confirmation Is a Separate Event

StockCharts describes the shooting star as a potential reversal that requires bearish confirmation. The exact confirmation rule still needs to be chosen. Different rules identify different event dates and produce different samples.

Rule A: Next close below the body

For a candidate at bar t, require bar t+1 to close below the lower edge of the shooting star body:

Close[t+1] < min(Open[t], Close[t])

This is an early confirmation rule. It does not require price to break the full candle low.

Rule B: Close below the shooting star low

Require a completed later candle to close below Low[t].

This is stricter and can confirm later. It excludes cases that move down from the body but never close below the full wick-to-wick range.

Rule C: Break plus follow-through

Require an initial break followed by a predefined additional decline within a fixed number of completed bars.

The project must define:

the break level;
whether a wick or close counts;
the additional movement threshold;
the number of bars allowed;
what happens if price first moves above the candidate high.

Rule D: Failed retest

Require price to move below a chosen reference, return to a predefined area, and then close lower under a written rejection or hold rule.

This identifies fewer and later events. It should not be combined with immediate confirmation results.

Rule E: No confirmation required

A project can study the raw candidate candle without confirmation. Label this version clearly. Do not compare it with confirmed versions while calling both “the shooting star strategy.”

Turn the Pattern Into a Testable Trade Rule

The phrase “trade the shooting star” is incomplete. A test needs more than a pattern name.

Freeze these fields before reviewing results:

Field	Example of what must be defined
Candidate rule	Exact prior-advance and candle-geometry version
Signal time	Pattern close, next close, later break, or retest event
Execution assumption	Next open, specified limit, stop trigger, or another rule
Invalidation	Candidate high, structural level, time expiry, or another written condition
Exit rule	Fixed horizon, opposite event, structural outcome, or another test rule
Cost model	Spread, commission, slippage, financing, and borrow constraints where relevant
Ambiguous-bar rule	What happens if an entry, stop, and target can occur in the same OHLC bar
Position rule	Whether overlapping signals are allowed and how exposure is limited
Sample split	Development sample, validation sample, and out-of-sample review

This table is a research specification, not a recommended trading plan. The values must match the market, data, execution model, and risk controls being tested.

OHLC data can hide event order

Suppose a later candle's high is above a hypothetical invalidation level and its low is below a hypothetical entry level. A four-price candle does not reveal which happened first.

Possible controls include:

use a smaller timeframe for event sequencing;
mark the bar ambiguous and exclude it;
apply a conservative worst-case assumption;
compare multiple intrabar assumptions;
avoid making fill claims the data cannot support.

This is one reason a clean chart result is not the same as live execution evidence.

Separate Classification, Confirmation, and Outcome

Use separate status fields.

Pattern classification

Not eligible: no qualifying prior advance or chart controls missing.
Candidate: closed candle meets the frozen definition.
Reclassified: later data correction or rule-version change alters the label.
Invalid definition: range is zero, body-ratio math is unstable, or required fields are missing.

Confirmation status

Unconfirmed: candidate exists but the chosen event has not occurred.
Confirmed: the predefined confirmation event occurred.
Expired: the confirmation window ended.
Disqualified: a predefined opposite event occurred first.
Ambiguous: OHLC data cannot establish event order.

Later outcome

moved lower under the chosen measurement rule;
moved higher first;
remained inside the candidate range;
crossed both directions in ambiguous order;
reached neither outcome before the horizon ended;
missing or invalid data.

A candidate that later fails is not necessarily a classification error. It may be a correctly identified pattern with an unfavorable outcome. Mixing those ideas inflates apparent accuracy because failed examples are quietly deleted.

Define Failure and Invalidation Before Testing

Possible failure rules include:

a close above the shooting star high before confirmation;
any trade above the high before confirmation;
no confirmation within a fixed number of bars;
a new confirmed higher high under the selected pivot rule;
continued closes above the candidate body;
a gap or data event that makes the execution assumption invalid.

These rules are alternatives, not universal requirements.

Also distinguish:

pattern invalidation: the original candle or context did not meet the written definition;
bearish-thesis failure: a valid candidate was followed by an opposite event;
trade invalidation: a hypothetical execution rule reached its risk condition;
data invalidation: the feed, adjustment, session, or timestamp changed.

Body Color: Record It, Do Not Overstate It

A shooting star can close above or below its open. Body color may be included as a variable, but it should not replace the geometry and context rules.

A controlled comparison can create two groups:

Close < Open
Close >= Open

Keep every other rule identical. Then compare frequency, confirmation, adverse movement, and outcome under the same market and horizon. Do not call one color stronger merely because a few selected charts look cleaner.

Volume and Indicators Are Optional Variables

Volume, momentum, moving averages, and calculated levels can be recorded, but they do not prove the pattern.

Volume controls

State whether the source is:

exchange volume;
consolidated volume;
futures volume;
broker-specific volume;
tick volume;
no usable volume.

Then define the baseline, such as a rolling median or average calculated without future data. “High volume” is not reproducible until the source and threshold are written.

A volume bar also does not identify institutions, retail traders, dealers, or algorithms.

Indicator controls

If RSI, MACD, a moving average, or another indicator is used:

record its formula and settings;
calculate it only from data available at the decision point;
treat it as a separate filter;
compare the candidate set with and without the filter;
preserve rejected examples.

Adding many filters can make historical examples look impressive while leaving too few observations to evaluate reliably.

Why There Is No Universal Shooting Star Win Rate

Published descriptions agree on the broad shape and the need for context or confirmation, but historical conclusions depend on definitions.

A statistical study of Shooting Star and Hammer patterns using decades of S&P 500 data reported materially different results depending on whether the outcome was measured with closing prices or intraperiod highs and lows. That is an important methodology lesson: changing the endpoint can change the conclusion even when the candidate candles are unchanged.

Results can also vary with:

market and instrument;
timeframe and session;
pattern geometry;
prior-advance definition;
confirmation rule;
measurement horizon;
dividends and contract adjustments;
transaction costs and shorting constraints;
development versus out-of-sample period;
treatment of overlapping and ambiguous events.

For that reason, this page does not use the previous unsupported 48%–64% win-rate claims, a 7/10 quality score, or a fixed best timeframe. A percentage without the dataset, rules, costs, sample dates, and outcome definition is not reusable evidence.

Worked Classification Example

Assume a completed candle has:

Open = 102.00
High = 108.00
Low = 101.80
Close = 102.40

Calculations:

Range = 108.00 - 101.80 = 6.20
Body = |102.40 - 102.00| = 0.40
Upper wick = 108.00 - 102.40 = 5.60
Lower wick = 102.00 - 101.80 = 0.20
Body share = 0.40 / 6.20 = 6.45%
Upper-wick/body ratio = 5.60 / 0.40 = 14.00

Under the example reference geometry, the candle qualifies. It becomes a shooting star candidate only if the project's prior-advance and chart-identity rules also pass.

The example does not yet say:

whether a later candle confirmed;
whether a hypothetical order could be filled;
whether the candidate preceded a decline;
whether the decline exceeded costs;
whether the result generalizes to another market.

Those questions belong to later fields in the record.

A Repeatable Shooting Star Checklist

Before labeling a candidate, answer each item:

Is the instrument and product identified?
Is the data provider or venue recorded?
Is the price type recorded?
Is the session and timezone fixed?
Is the timeframe fixed?
Is the adjustment method recorded?
Is the candidate candle closed?
Is Range > 0?
Is the geometry version named?
Does the body-size rule pass?
Does the upper-wick rule pass?
Does the lower-wick rule pass?
Does the body-location rule pass?
Is the near-zero-body rule handled?
Does the predefined prior advance pass?
Was the candidate available before future candles?
Is the location category recorded separately?
Is the confirmation version written?
Is the failure or expiry rule written?
Is the later measurement horizon fixed?
Are costs and execution assumptions recorded?
Are ambiguous bars preserved rather than forced into a result?
Is the candidate stored even if it later fails?
Is the test version locked before the next sample?

If a required field is missing, label the observation unresolved rather than filling it from hindsight.

Shooting Star Replay Worksheet

Use one row per candidate.

Chart identity

candidate ID;
instrument and product;
venue or provider;
price type;
session and timezone;
timeframe or bar type;
adjustment method;
volume source;
data date range.

Pattern definition

geometry version;
candle timestamp;
open;
high;
low;
close;
range;
body;
upper wick;
lower wick;
body share;
upper-wick/body ratio or alternative measure;
body-location measure;
body color group;
gravestone-doji tolerance result;
closed-bar confirmation.

Context

prior-advance version;
prior-advance result;
pivot or trend state;
raw resistance category;
calculated reference category;
range or breakout context;
higher-timeframe label if used;
event/news exclusion if predefined.

Confirmation and outcome

confirmation version;
confirmation timestamp;
confirmation status;
failure version;
failure timestamp;
expiry timestamp;
maximum favorable movement under the chosen measure;
maximum adverse movement under the chosen measure;
close after fixed horizons;
ambiguous-event flag;
cost assumptions;
missing-data flag;
notes written before reveal;
notes written after reveal.

How to Practice With ChartMini

ChartMini is best suited for lightweight candle-by-candle replay. It can help you hide future historical candles and apply a frozen worksheet.

A controlled session can use this sequence:

Choose one instrument, provider, session, and timeframe.
Write the shooting star geometry and prior-advance versions.
Write the confirmation, failure, and expiry rules.
Hide future candles.
Advance only through completed bars.
Record every eligible candidate, including unclear and failed cases.
Write the candidate fields before revealing more candles.
Reveal a fixed evaluation horizon.
Grade classification, confirmation, ambiguity, and outcome separately.
Freeze the development version.
Repeat on a later out-of-sample period.

You can begin a browser-based practice session at ChartMini.

What ChartMini does not reproduce

ChartMini does not reproduce:

live broker routing;
bid/ask queue position;
all spreads and commissions;
partial fills and rejections;
borrow availability or short-sale restrictions;
every financing and margin event;
exact intrabar order when only OHLC candles are available;
participant identity or hidden orders;
the emotional effect of live gains and losses.

Replay results are historical simulation records, not proof of future profitability.

Common Shooting Star Errors

Naming the shape without an advance

A long upper wick after a decline may be an inverted-hammer candidate. Inside a range, it may be only an upper-wick candle. Context is part of the label.

Using a live candle

A provisional upper wick can disappear before the interval closes.

Switching geometry versions

Using 2× body for losing examples and 3× body for winning examples makes the dataset unreproducible.

Dividing by a near-zero body

A doji-like candle can create an unstable wick/body ratio. Use a separate tolerance.

Drawing resistance after the reversal

The level must exist before later candles are revealed. Otherwise, location is selected from the result.

Treating confirmation as guaranteed follow-through

A later close below the candidate is an event under one rule. It is not proof that the decline will continue.

Inferring trapped buyers or institutional selling

OHLC data do not identify who traded, why they traded, or whether orders remain.

Using one fixed timeframe for every market

There is no universal minimum or best timeframe. Session structure, data quality, costs, and decision horizon differ.

Deleting failed or unresolved candidates

A study that keeps only clean reversals cannot estimate the frequency of failure, ambiguity, or no confirmation.

Treating a replay fill as a live fill

Historical candles do not reproduce every spread, queue, slippage, borrow, rejection, and liquidity condition.

Practical Next Step

Choose one market and one fixed chart configuration. Write one shooting star geometry version, one prior-advance rule, one confirmation rule, one failure rule, and one evaluation horizon. Record at least the candidate, confirmed, failed, expired, and ambiguous categories without changing the definitions. Only after the version is frozen should it be compared on a separate historical period.

Frequently Asked Questions

What is a shooting star candlestick?

A shooting star is a single-candle pattern considered after a defined price advance. It usually has a small real body near the lower part of the candle, a long upper wick, and little lower wick. The candle proves that price traded above the body and closed below its high; it does not by itself prove a bearish reversal.

What is the difference between a shooting star and an inverted hammer?

The geometry can be identical. A shooting star is classified after a predefined advance, while an inverted hammer is classified after a predefined decline. The prior-price context, not the candle color or upper-wick shape alone, determines which label is being tested.

Does the color of a shooting star candle matter?

Body color can be recorded, but it is not a universal requirement. A bullish-colored or bearish-colored body can meet the same geometric rule. If color is used as a filter, define it before reviewing results and compare it as a separate variable rather than changing the pattern definition afterward.

How should a shooting star be confirmed?

Confirmation must be defined before future candles are revealed. Possible rules include a later close below the shooting star body, a close below its low, a break followed by additional downside within a fixed horizon, or a failed retest. These rules create different datasets, so they should not be mixed in one test.

When is a shooting star invalidated?

A candidate can be invalidated if the candle was not closed, the required prior advance was absent, the geometry failed the written thresholds, or the chart data changed. After a valid candidate is recorded, later movement above its high can be labeled a failed bearish outcome under a predefined rule, but that outcome should remain separate from the original pattern classification.

Can ChartMini prove that the shooting star pattern is profitable?

No. ChartMini can hide future historical candles so you can classify candidates, apply fixed confirmation and failure rules, and review later price action. It does not reproduce every live spread, fill, slippage, fee, liquidity condition, margin event, or decision, and replay results cannot prove future profitability.

Sources and Evidence Boundaries

StockCharts ChartSchool — Introduction to Candlesticks: describes the Shooting Star as a potential bearish reversal after an advance, distinguishes it from the Inverted Hammer, and states that bearish confirmation is required.
CME Group — Technical Analysis: explains that candlestick charts display open, high, low, and close and that interpretation introduces subjectivity beyond the price data.
Fidelity Investments — Identifying Chart Patterns: includes the Shooting Star in its candlestick-pattern education and treats it as a one-candle pattern requiring context.
Journal of International & Interdisciplinary Business Research — Statistical Analysis of Japanese Candlesticks: illustrates how Shooting Star results can change with the selected price endpoint and methodology.
CFTC — Commodity Trading Systems Sold on the Internet: explains the inherent limitations of hypothetical and simulated trading results.

Shooting Star Candlestick: Definition, Confirmation, and Failure Rules