Wyckoff method in plain language
The Wyckoff method reads a market as a cycle of accumulation, markup, distribution and markdown. A single candle is not enough: the relationship between price spread, location and volume matters across a sequence. After a trend, price often enters a trading range where supply is absorbed or inventory is distributed. The scanner does not predict the next move. It organizes observable facts: where the prior trend stopped, which boundaries were tested and whether volume supports a breakout.
Accumulation and distribution
Accumulation commonly follows a decline. A selling climax and automatic rally establish the first range boundaries. Distribution is the mirror context after a rise, beginning with a buying climax and automatic reaction. A range alone is not a trade setup. Phase C provides the decisive test and phase D provides evidence of an exit. The scanner therefore shows structure separately from phase and does not turn every sideways market into a signal.
Wyckoff phases A–E
Phase A stops the previous trend and establishes boundaries. Phase B builds a cause through repeated movement inside the range. Phase C tests supply with a spring or demand with an upthrust after distribution. Phase D is a sign of strength or weakness, while phase E is the trend outside the old range. The labels are generated from ATR-normalized price movement and relative volume so one rule set can be applied across markets.
Spring, upthrust and volume
A spring briefly trades below support and closes back inside the range. An upthrust or UTAD performs the mirror test above resistance. The algorithm measures penetration in ATR units, the closing position and relative volume. A high-volume event remains awaiting confirmation until a later test occurs. Forming candles and future bars never confirm an event, preventing the most common repainting mistake in automated Wyckoff tools.
How to read the scanner
Choose an asset class and timeframe first. Phase C and D rows are sorted near the top because they are closest to a decisive test or exit. Click a row to see the same events on the candlestick chart. Range position shows where the latest close sits between support and resistance. Where the feed provides tick or proxy volume, the verdict says so directly and avoids pretending that it is centralized exchange volume.
Limits and risk
Wyckoff markup always contains interpretation. Real markets do not have to follow a textbook schematic, and two analysts can label the same range differently. Fixed ATR and relative-volume rules reduce subjectivity but cannot create certainty. Use the result as context, not as an order. Check liquidity, news, invalidation and acceptable risk before acting. A phase label is useful only when it agrees with a complete trading plan.
Effort and result
Effort refers to relative volume and candle spread; result means the actual closing move against a known range. Heavy volume without progress may be absorption, but it does not reveal intent on its own. The scanner marks that mismatch separately and waits for later closed bars before confirming a structure. Crypto volume belongs to a particular venue. Forex and continuous commodity contracts only provide a proxy; their readings cannot be compared as if they were centralized exchange volume. The panel explains that limit instead of using proxy data to claim confirmation of a spring or breakout.
Does the method work: outcome statistics
Wyckoff is useful as a disciplined description of market structure, not proof of a profitable strategy. The scanner checks each Spring and UTAD separately after the event: whether the range target is reached, the invalidation level is breached, or the target remains unreached after thirty subsequent closed bars. Until the window is complete the outcome is waiting. Instrument panels show cases over two years and median bars to a target, but suppress percentages below twenty completed cases. These rates omit executable entry prices, fees, slippage and liquidation risk. Evaluating a trading rule requires a separate blinded test, an out-of-sample period and controls for selection across markets.
Four structure schematics
These schematics illustrate sequence, not real prices or trade instructions.
Why the scanner waits for candle close
An unfinished candle can still change its extreme, close and volume. Naming it a spring too early would make a temporary break look permanent and then appear to repaint. Our calculation takes closed historical bars from the unified database and aggregates hourly data into 4H and 1D without reading the open interval. A phase transition also needs a causal sequence: climax and reaction establish A, the range survives at least eight bars in B, a decisive test establishes C, and a confirmed exit establishes D. An event and its later outcome are separate states, so failure does not erase an observed event. The screen shows Moscow time with a UTC tooltip to make each source bar auditable. Provider delay is a data-quality warning, never evidence of a new setup.
How range boundaries are formed
After a decline the algorithm looks for a selling climax and automatic rally; after a rise it looks for a buying climax and automatic reaction. These are candidates until a later secondary test confirms the structure. Boundaries come from extremes already observed, not from a future high that would make a historical chart look cleaner. If the expected reaction does not arrive inside the allowed window, the candidate expires. This matters on forex, where session pauses can resemble a neat sideways market without a robust trading range. Position is shown as a percentage only when the latest close is between boundaries. Outside the range the interface says above or below instead of displaying a misleading value such as 240 percent.
Phase B is cause, not a trade signal
In B, price explores both sides of the range while supply and demand rebalance. The algorithm cannot advance to C on a single interesting candle: A must be followed by at least eight completed B bars. Repeated boundary tests are expected, and a volume spike without proportional movement is shown as an effort-versus-result discrepancy. That discrepancy does not promise a reversal. Centralized exchange volume has a different meaning from forex tick volume and continuous-contract proxy volume, so proxy instruments are flagged and receive limited confirmation. A row with no confirmed range is a meaningful result rather than a hidden error. Empty timeframes for one instrument are collapsed in the all-timeframe view so they do not inflate visible counters.
Phase C and invalidation
A spring dips beneath support and returns to the range; UTAD tests resistance in a mirror structure. Penetration in ATR units, closing location and a later test all matter. A high-volume probe remains a candidate until a separate confirming bar. Spring and SOS cannot occur on the same candle, because testing supply and proving strength are different observations. If the test does not develop within fifteen bars, the structure returns to B instead of waiting forever. When price violates the invalidation level, history records an invalidated or failed outcome rather than relabeling old candles after the fact. This distinction lets readers separate an interesting shape from a confirmed causal sequence.
Phase D, exits and target checks
After C, SOS shows strength or SOW shows weakness outside the balancing area. An LPS or LPSY return may retest the former boundary, but it cannot guarantee trend continuation. E requires a later, distinct bar after the breakout; a jump across the creek must not be assigned to the same candle. Each decisive event is assessed on subsequent closed bars: the target was reached, the setup failed, or the target remained unreached after thirty bars. Before the window completes the outcome is waiting. Percentages are hidden for fewer than twenty completed observations because a small sample gives an unstable impression of accuracy. These outcomes describe historical markup, not net trade returns after commissions, financing and slippage.
Why timeframes are not interchangeable
The 1H view is available for crypto assets and major currency pairs, whose hourly histories are comparatively continuous. Metals, energy contracts and indices use 4H and 1D here so session gaps do not masquerade as micro-phases. Recent events over 24 hours are counted from the latest closed bars: twenty-four hourly bars, six four-hour bars or one daily bar. This definition is easier to reproduce than comparing a random browser refresh with a provider timestamp. Rows from different timeframes are distinct models: C on 4H does not invalidate B on 1D. Changing a filter alters the presentation and cannot retroactively create an event.
How to use event history
The history starts with thirty recent decisive events: Spring, Test, LPS, UTAD, SOS, SOW and LPSY. Show more reveals another batch without changing chronological order. Each entry carries an instrument, timeframe, confirmation state and outcome. Hovering a Moscow timestamp reveals UTC. Selecting a row opens the detailed chart for the same instrument, with its marker placed over the original closed candle. Earlier SC, AR and ST help construct boundaries but do not flood the list of decisive events. History is useful for auditing causal rules and identifying disputed cases, yet a sample selected from known events is not a substitute for a blinded strategy test.
Risk management and reproducibility
The scanner is an educational tool, not individualized investment advice. If a structure appears compelling, define the price action that would invalidate it before considering a position. Check instrument liquidity and the market session. A breakout can return into the old range; that is market behavior rather than proof that the model is broken. With leverage, even a correct directional view can lose money because of position size, fees, slippage or gaps. Do not translate event-outcome ratios into expected returns without testing actual orders separately. For a service audit, compare one closed bar in the screener, the home card and detailed SVG chart; a delayed source should trigger a freshness warning.
A causal walkthrough without a return promise
Imagine a market that declined for weeks and then prints a wide candle on heavy volume. That may be a selling climax, but the lower boundary remains a candidate until an automatic rally and secondary test arrive. After at least eight bars inside the range, a brief break below support could become a spring. Heavy volume still requires a later test; a confirmed SOS needs another distinct closed bar. E cannot occur on the breakout candle itself. The spring outcome remains waiting until subsequent candles hit a target, violate invalidation or complete the thirty-bar window. This sequence explains computation, not an instruction to buy. A neighboring daily timeframe may still be in B, and different markets have incomparable volume sources.
The scanner is informational. It does not issue buy or sell commands.
FAQ
What is the Wyckoff method?
It describes a sequence of price and volume behavior: a trend halt, a range, a boundary test and an exit. Accumulation, markup, distribution and markdown are cycle stages rather than trade commands. The scanner labels observable closed candles and confirmation conditions. It cannot know participant intent or guarantee that a market repeats a textbook pattern.
How is accumulation different from distribution?
Accumulation typically follows a decline and probes the lower range boundary. Distribution follows a rise and probes the upper boundary; a buying climax mirrors a selling climax. The scanner displays structure type separately from phase A–E because a range alone is not a decisive event. When confirmation conditions are missing, it explicitly reports no range.
What is a Wyckoff spring?
A spring briefly moves below range support but closes back inside. The algorithm also compares penetration with ATR, volume with its baseline and a later test. A heavy-volume probe is not confirmed on the same candle automatically. If subsequent closed bars breach the invalidation level, the observed event stays in history with a failed outcome.
Does Wyckoff work on crypto and forex?
The same ATR-normalized conditions can be applied to both, but volume provenance differs. A crypto venue reports its own executions; forex usually offers tick activity or another proxy. The panel labels that limitation and does not pretend proxy readings are centralized exchange volume. Scanner outcomes cannot be treated as expected trade returns without a separate execution-aware backtest.
Why does an event remain waiting?
An observed event and its outcome happen at different times. A spring can already be a confirmed observation while its target or invalidation has not occurred. Later closed bars are needed, over a window of at most thirty bars on the selected timeframe. Until that window resolves, the service says waiting and never inserts future information into an earlier candle.
What does phase E mean?
Phase E means price has left the former range after a confirmed exit and a separate later candle. The old area remains visible for a limited time as structural context. E cannot be assigned to the same bar as SOS or SOW, because that would confirm a later state prematurely. It describes algorithm state, not the probability of trend continuation.
Why are percentages hidden for small samples?
If fewer than twenty Spring or UTAD cases have completed for an instrument and timeframe, one new observation can change the target rate sharply. The panel therefore shows the sample count while suppressing percentages and the median. Even with larger samples these are historical markup outcomes, not net returns from executed orders after fees, slippage and risk.