RSI Divergence + EMA Trend FilterDescription:
Divergence is one of the most discussed concepts in technical analysis and one of the most misapplied. The core idea is simple: when price makes a new high but the RSI makes a lower high, momentum is weakening even as price advances. That disconnect between price action and momentum is what divergence measures — and it often precedes reversals before price itself confirms the change in direction.
This strategy formalizes that concept into a rule-based, backtestable system with two components: RSI divergence detection and an EMA trend filter that determines which divergences to act on.
What divergence actually measures
RSI measures the speed and magnitude of price changes. When price reaches a new swing high but RSI fails to reach a correspondingly higher reading, it means the buying pressure behind the new high was weaker than the buying pressure behind the previous high. The market got to a higher price but required less momentum to do it — which suggests the move is losing conviction. Bearish divergence (price higher, RSI lower) signals potential exhaustion in an uptrend. Bullish divergence (price lower, RSI higher) signals potential exhaustion in a downtrend.
Important: divergence is a momentum signal, not a reversal guarantee. Price can continue making new highs with weakening RSI for a significant period before actually reversing. This is why divergence signals work best when combined with a trend filter that identifies the broader market context.
The EMA filter
The 200 EMA defines the dominant trend regime. Bearish divergence signals — where momentum is weakening on the upside — are only acted on when price is below the 200 EMA, meaning the broader trend is already bearish and divergence represents a potential resumption of that trend after a counter-trend bounce. Bullish divergence signals are only acted on when price is above the 200 EMA, where they represent potential continuations of the dominant uptrend after a pullback with improving momentum.
This filter deliberately reduces the total number of signals. Many valid divergences occur against the dominant trend and produce short-lived reversals that reverse again quickly. By requiring trend alignment, the strategy trades fewer setups but acts on the ones with a higher probability of following through.
How divergence is detected
The strategy identifies swing highs and swing lows using a lookback period — the number of bars on each side of a pivot that must be lower (for a high) or higher (for a low) to qualify as a genuine swing point. When two consecutive swing highs show price making a higher high but RSI making a lower high, bearish divergence is flagged. When two consecutive swing lows show price making a lower low but RSI making a higher low, bullish divergence is flagged.
The lookback length is the most important input to tune. A shorter lookback detects more swing points and generates more signals, but many will be minor pivots in the context of noise. A longer lookback requires more significant swing points and generates fewer, higher-quality signals. On daily charts, a lookback of 5 works well. On lower timeframes, 3 to 4 is more appropriate.
Exits
Positions exit at an ATR-based stop-loss and a fixed ATR-based take-profit. The stop is placed beyond the swing point that generated the divergence signal — for a bearish divergence, the stop sits above the swing high; for a bullish divergence, below the swing low. This is intentional: if price breaks through the very level that defined the divergence, the signal is invalidated regardless of what RSI was doing. The take-profit is set at 2x ATR to maintain a positive reward-to-risk ratio across the system.
What to evaluate in backtesting
Look at the signal distribution across different market environments. Divergence strategies tend to perform differently in trending versus ranging markets — in strong trending environments, bearish divergences against the dominant trend will produce many false signals even with the EMA filter. Look at whether the EMA filter is doing real work by temporarily disabling it and comparing signal quality. Check average trade duration — divergence signals that take too long to play out often give back open profit before the take-profit level is reached.
This is not a high-frequency strategy. On daily charts with a 5-bar lookback, signals may appear only a few times per month on a given instrument. That frequency is appropriate — divergence setups require specific conditions to form and should not be forced.
Shared for educational purposes and discussion. This is not investment advice. Backtest on your own instruments and timeframes before drawing conclusions about expected performance. 策略
[GYTS-CE] Kinetic Trend Envelope (adaptive trailing stop)Kinetic Trend Envelope (Community Edition)
🌸 Part of GoemonYae Trading System (GYTS) 🌸
🌸 --------- INTRODUCTION --------- 🌸
💮 What is the Kinetic Trend Envelope?
The Kinetic Trend Envelope (KTE) is an adaptive directional trailing stop in the lineage of SuperTrend, rebuilt around the premise that volatility is kinetic energy . It measures per-bar motion with five academically grounded volatility estimators, then widens the envelope as energy rises and contracts it as motion settles.
In an uptrend, the lower band ratchets higher and never retreats; in a downtrend, the upper band ratchets lower. The direction changes when the active stop is breached, after which the opposite side becomes the new trailing stop.
💮 Why Use This Indicator?
Conventional trailing stops typically combine a price anchor with one symmetric ATR-derived width. The KTE extends that model with:
Asymmetric volatility profiling — Bullish- and bearish-candle volatility shape the upper and lower bands independently.
Three direction-switch methods — High/low, close, or a smoothed estimator controls flip sensitivity without moving the band anchor.
Five volatility estimators — ATR plus Parkinson, Garman-Klass, Rogers-Satchell, and Yang-Zhang covers different treatments of gaps, drift, and intrabar range.
The outputs are calibrated to a common width basis, so Volatility Factor remains interpretable across estimators and price scales. Fine adjustment may still be useful, but switching estimators should not require re-tuning by orders of magnitude.
↑ The KTE on a trending instrument. The thick line is the active trailing stop; the thin line shows the opposing side of the envelope. Both expand and contract with market energy.
↑ KTE beside TradingView's built-in SuperTrend, both using ATR with a 10-bar lookback. KTE's asymmetric profile changes how each side responds to directional volatility while the monotonic active band avoids premature loosening.
🌸 --------- HOW IT WORKS --------- 🌸
💮 Core Concept
The bands share a smoothed price estimator as their anchor, but use separate volatility profiles:
Upper band = estimator + (factor × bullish-candle volatility)
Lower band = estimator − (factor × bearish-candle volatility)
In a bullish state, the lower band is active and can only rise. In a bearish state, the upper band is active and can only fall. This monotonic constraint prevents a live trailing stop from loosening within the trend.
The selected direction-switch method changes only the breach test. It does not change the smoothed estimator anchoring the envelope, so a wick-sensitive trigger cannot drag the bands around with the wick.
💮 The Five Volatility Estimators
Each estimator reads a different part of the OHLC bar:
ATR (Wilder, 1978) — Familiar baseline that handles gaps through true range.
Parkinson (1980) — Uses high-low range; efficient under continuous, low-drift conditions.
Garman-Klass (1980) — Adds open-close information; favours continuous sessions without material gaps.
Rogers-Satchell (1991) — Drift-independent and well suited to trending, continuously traded instruments.
Yang-Zhang (2000) — Combines overnight gaps, open-close movement, and Rogers-Satchell; the gap-aware default.
Statistical efficiency does not guarantee a visibly tighter stop. At slow Adaptation Speed settings, long averaging makes the estimators look similar; at fast settings, their different treatments of gaps, drift, and range become more visible. Choose according to the instrument's behaviour rather than expecting one estimator always to produce the narrowest band.
↑ ATR and Yang-Zhang at Adaptation Speed 2. The long profile memory (low speed) smooths away most of the difference, so the two envelopes nearly overlap.
↑ ATR and Yang-Zhang at Adaptation Speed 8. The short profile memory (high speed) exposes their different volatility readings, producing visibly distinct envelope widths.
💮 Asymmetric Volatility Profiling and Adaptation Speed
The KTE stores volatility from bullish and bearish candles separately. Bullish samples determine the upper width; bearish samples determine the lower width. This allows the two sides to respond differently when upward and downward motion carry different energy.
Adaptation Speed controls the memory of this profile, not the speed of the price estimator and not the distance of the stop by itself. Its 1–10 scale maps logarithmically to an internal window:
Speed 3 — approximately 878 bars: stable and slow to re-weight
Default 3.5 — approximately 570 bars: general-purpose smoothing
Speed 8 — approximately 11 bars: highly responsive to recent volatility
Speed 10 — approximately 2 bars: extremely reactive and noisy
Faster does not necessarily mean closer to price. During a volatility burst, a fast profile recognises the expansion sooner and may widen the band sharply. Because the active stop cannot loosen, it can then remain flat until the estimator catches up. A slow profile dilutes the same burst across much more history, so its narrower band may appear to follow price faster.
This is why two instances matched during a calm period can separate during a shock, especially when they also use different Volatility Factor values. Compare Adaptation Speed with the same factor first; matching lines in one regime does not make two configurations equivalent elsewhere.
The profiles are also direction-conditioned: bullish samples are replaced by later bullish candles and bearish samples by later bearish candles. A recent high-volatility sample can therefore persist through a run of opposite-colour candles, producing deliberate step-like plateaux in the relevant band.
↑ Asymmetric profiling in action: the upper and lower widths respond independently to bullish- and bearish-candle volatility.
💮 Direction Switch Methods
The breach source sets the balance between responsiveness and false flips:
On high/low — Uses the current bar's wick and can switch on the breach bar. Fastest and most sensitive to noise.
On close — Uses the previous confirmed close; the switch appears on the following bar.
On estimator — Uses the previous smoothed estimator; the most conservative default, also switching on the following bar.
↑ The three switch methods share the same band geometry but change direction at different times.
🌸 --------- KEY FEATURES --------- 🌸
💮 Eight Estimator Filters
The configurable price anchor includes:
Ultimate Smoother, 2- or 3-pole — Low-noise, near-zero-lag passband response; the 2-pole version is the default.
Super Smoother, 2- or 3-pole — Ehlers low-pass filters for progressively stronger smoothing.
BiQuad — Second-order low-pass filter with an adjustable Q-factor.
ADXvma — Adapts to trend strength and tends to flatten in ranges.
MAMA — Cycle-adaptive MESA moving average.
A2RMA — Adaptive recursive moving average with adjustable gamma.
They are provided by the open-source FiltersToolkit library.
💮 Visual Layering
The display separates function from context:
Active band — Thick directional trailing-stop line
Opposing band — Thin reference for the inactive side
Channel fill — Visual separation between the estimator and each band
Estimator — Optional smoothed anchor
Palette, light/dark mode, widths, and transparencies can be adjusted independently.
🌸 --------- USAGE GUIDE --------- 🌸
💮 Getting Started
Start with the defaults, observe several calm and volatile regimes, and change one dimension at a time:
Tune Volatility Factor for the preferred stop distance.
Tune Adaptation Speed for how quickly width should respond to regime changes.
Choose the direction-switch method for the preferred confirmation level.
Change the volatility estimator only when its assumptions better fit the instrument.
💮 Choosing a Volatility Estimator
Gapped equities — Yang-Zhang accounts for overnight movement.
Trending 24/7 markets — Rogers-Satchell is drift-independent without a separate gap component.
Continuous, range-led markets — Parkinson or Garman-Klass offers efficient range-based measurement under their assumptions.
Familiar baseline — ATR provides conventional true-range behaviour.
On continuous instruments, Rogers-Satchell and Yang-Zhang may look very similar because there are few gaps to distinguish them. Use the Volatility Toolkit to compare their raw behaviour on the intended instrument.
↑ Three estimators compared on one instrument, each reading a different combination of OHLC information.
💮 Tuning Width and Responsiveness
These controls solve different problems:
Volatility Factor — Sets the distance per unit of measured volatility.
Adaptation Speed — Sets the memory of the bullish/bearish profile; faster can widen the stop sooner during shocks.
Volatility Lookback — Sets how quickly the underlying per-bar volatility estimate changes.
Estimator Lookback — Sets the smoothness of the price anchor.
Use symptoms to guide adjustment:
Frequent flips on minor pullbacks — Increase Volatility Factor or use a more conservative switch method (e.g. "on estimator").
Excessive give-back — Decrease Volatility Factor or use a more responsive switch method (e.g. "on high/low").
Width reacts too slowly to regime changes — Increase Adaptation Speed or reduce Volatility Lookback.
Bands become erratic during shocks — Reduce Adaptation Speed or increase Volatility Lookback.
↑ A tight factor follows price more closely and flips more often; a loose factor tolerates larger pullbacks.
💮 Trading Applications
Discretionary trailing stop — Move a protective stop with the active band as it tightens.
Trend confirmation — Accept long signals only during a bullish KTE state, and short signals only while bearish.
Exit timing — Treat a direction change as an exit when the trade thesis is trend-following.
💮 Integration with GYTS Suite
The visible bands and estimator can be selected as sources by compatible Pine scripts. Two packed streams are also exposed:
🔗 STREAM KTE 🪜 Trailing Stoploss — Positive lower-band value in a bullish state; negative upper-band value in a bearish state.
🔗 STREAM KTE 🪜 Mechanism — Encodes the switch method and scale-invariant estimator relationship for compatible consumers.
The KTE is, first and foremost, a trailing stop, and these streams are built for stop management. The Order Orchestrator strategy consumes the Trailing Stoploss and Mechanism streams together : the first supplies the active stop level and its direction, the second makes the strategy's trailing-exit runner follow whatever switch method and estimator you set here. So the stop is configured once, in the KTE.
Beyond that primary role, the signed trailing-stop stream can also serve as a trend signal, since its sign flips with direction: it can be read through sign and magnitude as an entry/exit signal, including by Flux Composer . The KTE can also be paired with Market Regime Detector so flips are acted on only when the broader regime supports trend-following behaviour.
🌸 --------- LIMITATIONS --------- 🌸
Trailing-stop latency — Every trailing stop gives back some of the move between the trend extreme and the eventual breach.
Whipsaws in ranges — Low-energy chop can produce repeated flips; a regime filter may help when ranging conditions dominate.
Fast adaptation can widen the stop — Higher Adaptation Speed means faster volatility response, not guaranteed proximity to price.
Direction-conditioned memory — A bullish or bearish outlier remains in its own profile until enough matching-direction samples replace it, which can create plateaux after shocks.
Warm-up and sample size — Long profile windows need sufficient chart history; strongly one-sided markets may leave one side with few recent samples.
🌸 --------- CREDITS --------- 🌸
💮 Academic Sources
Wilder, J. W. (1978). New Concepts in Technical Trading Systems . Trend Research.
Parkinson, M. (1980). The Extreme Value Method for Estimating the Variance of the Rate of Return. Journal of Business, 53 (1), 61–65. DOI
Garman, M. B., & Klass, M. J. (1980). On the Estimation of Security Price Volatilities from Historical Data. Journal of Business, 53 (1), 67–78. DOI
Rogers, L. C. G., & Satchell, S. E. (1991). Estimating Variance from High, Low and Closing Prices. Annals of Applied Probability, 1 (4), 504–512. DOI
Yang, D., & Zhang, Q. (2000). Drift-Independent Volatility Estimation Based on High, Low, Open, and Close Prices. Journal of Business, 73 (3), 477–491. DOI
Ehlers, J. F. (2024). The Ultimate Smoother. Technical Analysis of Stocks & Commodities , 2024-04. TASC
Ehlers, J. F. (2004). Cybernetic Analysis for Stocks and Futures . Wiley. Covers SuperSmoother, MAMA and more.
💮 Inspiration
Thanks to Trendoscope for inspiring us with the Supertrend - Ladder ATR (2021). It derives long-side stop distance from bearish-candle ATR and short-side distance from bullish-candle ATR, which is one of the mechanisms that we tried to develop further with the KTE.
💮 Libraries Used
FiltersToolkit — Ultimate Smoother, Super Smoother, BiQuad, ADXvma, MAMA, and A2RMA
VolatilityToolkit — Parkinson, Garman-Klass, Rogers-Satchell, and Yang-Zhang estimators
MathTransform — Logarithmic scaling for Adaptation Speed
ColourUtilities — Palette management and light/dark-mode colour adjustment
指标
KC Sessions PRO [Asia, London & New York High/Low]Description
KC Sessions PRO is a clean session-structure indicator designed to display the active trading ranges of the Asia, London, and New York sessions.
The indicator tracks each enabled session’s developing high and low and presents the information using lightweight session boxes, high/low lines, compact labels, and a dashboard. Its main objective is to provide session context while keeping the chart readable and reducing unnecessary historical clutter.
HOW IT WORKS
During an active session, the script continuously updates that session’s high and low.
When the session closes, its completed range can either be removed or retained temporarily, depending on the selected display mode.
DISPLAY MODES
• Clean — Displays active sessions only and removes completed session objects.
• Standard — Displays active sessions and retains limited recent session history.
• Pro — Displays active sessions with additional completed-session context.
MAIN FEATURES
• Asia, London, and New York session tracking
• Developing session high and low
• Optional session range boxes
• Solid session-high lines
• Dashed session-low lines
• Compact A, L, and NY labels
• Clean, Standard, and Pro display modes
• Automatic removal of older chart objects
• Adjustable completed-level extension
• Configurable session time zone and hours
• Compact active-session dashboard
• Session open and close alert conditions
DASHBOARD
The dashboard displays:
• Current active session
• Active session high
• Active session low
• Selected display mode
• Timeframe suitability status
DEFAULT SESSION HOURS
The default hours use the selected time zone:
• Asia: 00:00–08:00
• London: 08:00–16:00
• New York: 13:00–21:00
Users should adjust these hours and the time zone where necessary for their instrument, broker, location, and daylight-saving requirements.
RECOMMENDED USE
This indicator is most useful on intraday charts, particularly:
• 1-minute
• 5-minute
• 15-minute
• 30-minute
• 1-hour
• 4-hour
PRACTICAL APPLICATIONS
KC Sessions PRO may be used to study:
• Session range expansion
• Intraday support and resistance
• Breaks of session highs or lows
• Volatility changes between trading sessions
• London and New York overlap
• Session-based market structure
• Intraday liquidity context
IMPORTANT LIMITATIONS
• Session times depend on the selected time zone and user-defined session hours.
• Different brokers and instruments may use different trading-day structures.
• The displayed session high and low continue developing until the session closes.
• Session levels should not be treated as automatic trade-entry signals.
• The script does not predict future price direction or guarantee trading outcomes.
This indicator is intended for educational and informational analysis only. It is not financial advice, a trading recommendation, or a guarantee of performance. Users should conduct their own analysis and apply appropriate risk management.
Short Description:
A clean intraday indicator that tracks Asia, London, and New York session ranges, developing highs/lows, limited session history, and active-session context.
Release Notes
INITIAL RELEASE — VERSION 2.1
• Added Asia, London, and New York session tracking
• Added developing session highs and lows
• Added Clean, Standard, and Pro display modes
• Added session boxes and compact session labels
• Added automatic historical-object cleanup
• Added active-session dashboard
• Added configurable timezone and session hours
• Added session open and close alert conditions
• Optimized default settings for a cleaner chart
指标
指标
指标
Trading Context Panel + Auto BiasTrading Context Panel + Auto BIAS
Trading Context Panel + Auto BIAS is a Pine Script v6 indicator that displays an operational context panel directly on the chart.
It allows the user to record and view the market context in HTF, MTF, and LTF, analysis date/time, manual trend direction, automatic BIAS, and a custom narrative for each registered asset.
The indicator supports up to 4 different assets. When switching charts, it automatically displays only the information corresponding to the current asset.
If the current chart asset is not registered, the panel will display:
No context for this asset
Main Configuration
Inside each section Asset 1, Asset 2, Asset 3, and Asset 4, the user can configure:
- Asset: asset/pair name
- Analysis date: analysis date
- Analysis time: analysis time
- HTF timeframe: HTF timeframe
- HTF Trend: BULLISH, BEARISH, or SIDEWAYS
- MTF timeframe: MTF timeframe
- MTF Trend: BULLISH, BEARISH, or SIDEWAYS
- LTF timeframe: LTF timeframe
- LTF Trend: BULLISH, BEARISH, or SIDEWAYS
- Context: free text field for the analysis narrative
Automatic BIAS
The indicator includes an automatic BIAS column for the HTF, MTF, and LTF defined by the user.
The BIAS is calculated using ICT-style liquidity sweep logic, based on the previous candle high/low and the candle close.
Basic logic:
- Sweeps the previous high and closes above it = bullish bias
- Sweeps the previous low and closes below it = bearish bias
- Sweeps the previous high and closes back below it = bearish rejection
- Sweeps the previous low and closes back above it = bullish rejection
- Sweeps both sides = the candle direction defines the bias
This allows the trader to compare:
- Trend: manual trend defined by the user
- BIAS: automatic reading generated by the script
Previous Highs and Lows
The script can display important previous high/low levels:
- PDH — Previous Day High
- PDL — Previous Day Low
- PWH — Previous Week High
- PWL — Previous Week Low
Each level can be shown or hidden individually.
The script also includes UTC / Timezone configuration for previous day/week calculations.
Breack Levels and Premium / Discount
The indicator can optionally display:
- Breack Levels based on the sweep of the previous candle high/low
- Premium / Discount zones on the candle that generated the sweep
These features can be enabled or disabled in the settings.
Important: Asset Name
The Asset field is essential.
The typed name must match the asset name displayed on the TradingView chart.
Examples:
- If the chart shows EURUSD, enter EURUSD
- If the chart shows GBPUSD, enter GBPUSD
- If it only shows XAUUSD, enter XAUUSD
If the asset name does not match the current chart, the indicator will not find the context and will display:
No context for this asset
Visual Settings
The user can customize:
- show/hide panel
- panel position
- text size
- context width
- background color
- text color
- border color
- show/hide HTF, MTF, and LTF
- show/hide asset name
- show/hide analysis date and time
- show/hide PDH, PDL, PWH, and PWL
- show/hide Breack Levels
- show/hide Premium / Discount zones 指标
Market Regime Engine [NQ Labs] WHAT THIS ANSWERS
Almost every indicator tells you WHAT the market is doing. Very few tell you
whether the current market has enough exploitable structure to be worth
risking money on at all. The Market Regime Engine is built for that second
question: should I be trading this right now, and if so, what style of
approach do current conditions actually favour?
It is not a signal generator. It produces no entries. It is a filter you run
underneath whatever you already trade.
WHY IT IS BUILT THIS WAY
Two design decisions separate this from a conventional regime or trend filter.
1. EVERY AXIS IS PERCENTILE-RANKED AGAINST ITS OWN HISTORY.
Most regime tools use fixed thresholds - an ADX of 25, an ATR multiplier of 3.
Those numbers are calibrated to whatever the author tested on. They are the
reason a tool behaves sensibly on one instrument and nonsensically on the
next. Here, each axis is ranked against its own recent distribution on the
current symbol and timeframe, so a 65th-percentile reading carries the same
meaning on a crypto pair as it does on an FX major. Nothing needs retuning
per market.
2. THE THREE AXES ARE ORTHOGONAL BY DESIGN.
Many "confluence" tools stack three views of momentum and present the
agreement as confirmation. Correlated inputs agreeing is not evidence. These
three measure genuinely different properties:
DIRECTIONAL EFFICIENCY - net distance travelled divided by the total path
walked to get there. A value near 1 is a straight line. A value near 0 means
price ended where it started after a great deal of work. This is the most
direct available answer to "is there a trend here", and it is independent of
direction.
VOLATILITY STATE - ATR ranked against its own history. Not "is volatility
high" in absolute terms, but "is volatility high for this instrument".
RETURN PERSISTENCE - lag-1 autocorrelation of log returns. Positive means an
up bar tends to be followed by an up bar, so momentum approaches have
something to work with. Negative means the market reverts. Near zero means
returns carry no usable memory.
THE SCORING LOGIC
The Tradeability Score (0-100) rests on one idea:
Efficiency is the signal. Volatility is the cost of being wrong.
Volatility only hurts you when there is no direction to exploit.
So the noise penalty is the product of how little direction exists and how
violent the market is. It approaches zero inside a strong trend no matter how
wild conditions get - because volatility inside a trend is opportunity, not
noise - and it peaks precisely where traders are most reliably ground down:
low efficiency combined with high volatility. That specific combination is
what the engine exists to flag.
THE FIVE REGIMES
TREND Efficiency high. Directional. Follow it.
VOLATILE CHOP Efficiency low, volatility high. No direction, maximum cost
of being wrong. The state that does the damage.
COIL Efficiency low, volatility low. Compression. Wait.
RANGE Efficiency low, volatility middling. Edges are tradeable.
TRANSITION Efficiency mid-range. No clean read. Reduce size.
A raw regime read must hold for a configurable number of consecutive bars
before it is confirmed. Without that filter, readings that straddle a
threshold flip back and forth for single bars.
REGIME MATURITY - THE PART I HAVE NOT SEEN ELSEWHERE
The engine records the length of every completed regime run on the current
chart and reports the current run against that learned baseline, together with
the sample size behind it.
A trend at 0.3x its typical duration and a trend at 2.4x are not the same
trade, even though every conventional indicator prints the identical reading
for both. A compression that has lasted three times longer than compressions
normally last on this instrument is a different proposition to one that
started six bars ago.
The sample size (n=) is shown deliberately. Below the configured threshold the
cell dims, because a baseline built from two observations is not evidence and
should not be dressed up as though it were.
HOW TO USE IT
The engine is a gate, not a trigger. The intended workflow:
1. Read the Verdict first. STAND DOWN means your edge, whatever it is, is
probably not present. The most valuable output of this tool is the trades
it talks you out of.
2. Read the Regime and Favoured Style. Trend-following systems belong in
TREND. Mean-reversion belongs in RANGE, and only when Persistence confirms
the market is actually reverting rather than merely quiet. COIL means the
setup has not arrived yet. VOLATILE CHOP means nothing you own works here.
3. Read Persistence before choosing an approach. If it reads Random, the
market has no memory to exploit in either direction, and both momentum and
mean-reversion are coin flips.
4. Use vs Typical for position management, not entry. An extended regime is
not a reversal signal. It is a reason to stop adding and to tighten what
you already hold.
5. Let the ribbon and score do the work on the price chart. The dashboard can
be forced onto the main pane so the engine occupies a pane without costing
you chart real estate.
SETTINGS THAT ACTUALLY MATTER
RANKING LOOKBACK is the important one. It must span several complete regime
cycles on your timeframe or every axis will rank near the middle and
everything will read TRANSITION. 250 bars on a daily chart is roughly a year -
a real baseline. 250 bars on a 5-minute chart is under a day, and is not.
On intraday timeframes, raise it substantially.
MINIMUM REGIME RUN trades responsiveness for stability. Higher values give a
calmer read and a cleaner duration baseline at the cost of confirming changes
later. Set it to 1 to see the unfiltered classification.
CHOP PENALTY controls how hard low-efficiency volatility is punished. Set it
to 0 and the score becomes pure efficiency rank.
EFFICIENCY and VOLATILITY thresholds are percentiles, not raw values. They do
not need adjusting per instrument. That is the point.
LIMITATIONS - STATED PLAINLY
- This is a lagging, descriptive tool. It classifies conditions that have
already formed. It does not forecast, and it cannot tell you a regime is
about to end.
- Confirmation costs lag. A genuine regime change registers a few bars late by
design. That is the price of not flickering.
- The maturity baseline is learned from visible chart history and rebuilds
whenever settings, symbol, or timeframe change. It is not persistent memory.
Early in a chart, or for a rarely-occurring regime, the sample is thin. The
n= value tells you when to discount it.
- On intraday equity charts, overnight gaps inflate ATR and distort the
persistence reading. Daily and above, or extended hours, handle this better.
- Autocorrelation is a weak statistic on short samples. Persistence readings
close to zero should be read as "no information", not as a subtle signal.
- Efficiency is directionless. TREND tells you a trend exists, not which way
it points. Pair it with something that reads direction.
- No regime classification is correct at the moment it matters most, which is
the turn. Nothing here changes that.
This script is open-source. Read the code, disagree with it, improve it.
This is an analytical tool for studying market conditions. It is not financial
advice, and it does not generate trade recommendations. Trading involves risk
of loss. 指标
Forza Relativa - Professional Table v2█ OVERVIEW
Questo indicatore calcola e mostra la Forza Relativa (Relative Strength) di un asset rispetto a un benchmark di riferimento e al suo settore di appartenenza (se applicabile). Tutti i calcoli sono basati su giorni solari reali ("Calendar Days"), garantendo un confronto preciso e uniforme delle performance su diversi archi temporali: 1g, 7g, 30g, 60g, 90g e 180g.
A differenza delle tabelle standard, questa versione è stata interamente ottimizzata per il Tema Scuro (Dark Theme) di TradingView, garantendo una leggibilità nitida e professionale in ogni condizione di mercato.
█ CARATTERISTICHE PRINCIPALI
• Adattabilità Asset & Crypto: Lo script rileva automaticamente la classe di asset sul grafico. Se l'asset è una Criptovaluta, imposta come benchmark predefinito Bitcoin (BTC) o Ethereum (ETH) a scelta dell'utente. Se l'asset è un'Azione o ETF, utilizza lo S&P500 (SPY) come benchmark e attiva la colonna dedicata al "Settore".
• Selezione del Settore: Per il mercato azionario, è possibile selezionare manualmente l'ETF settoriale di riferimento (es. XLK per il Tech, XLF per i Finanziari, ecc.) per monitorare se l'asset sta sovraperformando o sottoperformando il proprio settore di appartenenza.
• Calcolo a Giorni Solari (Calendar Days): Le variazioni percentuali non si basano sul semplice conteggio delle candele del grafico corrente (che variano a seconda del timeframe), ma calcolano l'effettiva variazione temporale indietro nel tempo, rendendo l'indicatore ideale sia su grafici Daily che Intraday.
█ OTTIMIZZAZIONE GRAFICA & NITIDEZZA (DARK THEME)
Per risolvere il classico problema di scarsa leggibilità delle tabelle TradingView su sfondi scuri, sono state introdotte le seguenti migliorie:
1. Contrasto Dinamico: Il testo delle celle utilizza una palette di colori e trasparenze che si adatta allo sfondo per risaltare chiaramente senza affaticare la vista.
2. Palette Colori Soft: Sostituiti i verdi e rossi standard (troppo accesi e fastidiosi su sfondo scuro) con un elegante Verde Smeraldo Soft per i rendimenti positivi e un Rosso Corallo Soft per quelli negativi.
3. Griglia Protetta: La tabella presenta bordi grigio-scuri solidi per separare nettamente i dati e le colonne, rendendo l'interfaccia pulita e moderna.
█ PARAMETRI PERSONALIZZABILI (INPUTS)
• Crypto Benchmark: Consente di scegliere tra BTCUSDT ed ETHUSDT quando si analizzano le criptovalute.
• Dimensione Testo Tabella: Modifica la dimensione della tabella per adattarla a schermi di diverse risoluzioni (Tiny, Small, Normal, Large, Huge).
• Posizione Tabella: Permette di posizionare la griglia in qualsiasi angolo del grafico (es. in basso a destra, in alto a sinistra, ecc.) per non coprire i prezzi.
• Settore Manuale: Consente di selezionare l'ETF settoriale SPDR di riferimento per l'asset azionario sotto analisi.
█ NOTE SUL TIMEFRAME
Dato che l'indicatore analizza performance storiche fino a 180 giorni solari fa, si consiglia di utilizzare timeframe come il Giornaliero (1D) o l'Orario (1H/4H). Su timeframe estremamente ridotti (es. 1 o 5 minuti), TradingView potrebbe non avere abbastanza candele storiche caricate per calcolare i periodi più lunghi, mostrando la dicitura "n/a" (non disponibile).
Relative Strength vs Benchmark & Sector (Dark Theme)
█ OVERVIEW
This indicator measures and displays the Relative Strength (RS) of an asset by comparing its performance against a selected benchmark and, when applicable, its sector benchmark.
All calculations are based on actual Calendar Days rather than a simple bar count, ensuring consistent and accurate performance comparisons across the following periods:
1D, 7D, 30D, 60D, 90D, and 180D
The table has been specifically designed and optimized for the TradingView Dark Theme, providing maximum readability, strong contrast, and a professional appearance under all market conditions.
█ KEY FEATURES
• Automatic Asset Detection
The script automatically identifies the type of asset displayed on the chart:
Cryptocurrencies: uses either Bitcoin (BTC) or Ethereum (ETH) as the benchmark, selectable by the user.
Stocks and ETFs: uses SPY (S&P 500 ETF) as the benchmark and automatically enables the Sector comparison column.
• Sector Comparison
For equities, users can manually select the appropriate SPDR sector ETF (XLK, XLF, XLE, XLV, etc.) to determine whether the stock is:
Outperforming its sector
Underperforming its sector
Moving in line with its sector
• Calendar-Day Performance Calculation
Performance measurements are independent of the chart timeframe.
Instead of relying on the number of bars displayed, percentage changes are calculated using actual calendar days, ensuring consistent results across both Daily and Intraday charts.
█ DARK THEME VISUAL OPTIMIZATION
To improve readability on dark backgrounds, several visual enhancements have been implemented:
1. Dynamic Contrast
Text colors and transparency levels have been carefully tuned to provide excellent visibility while minimizing eye strain.
2. Soft Color Palette
Standard colors have been replaced with more balanced tones:
Soft Emerald Green for positive values
Soft Coral Red for negative values
This creates a cleaner appearance and improves readability during extended analysis sessions.
3. High-Definition Grid
The table uses solid dark-gray borders to clearly separate cells, columns, and headers, resulting in a modern and professional layout.
█ CUSTOMIZABLE INPUTS
• Crypto Benchmark
Select the benchmark used for cryptocurrency analysis:
BTCUSDT
ETHUSDT
• Table Text Size
Adjust the table size to fit different screen resolutions:
Tiny
Small
Normal
Large
Huge
• Table Position
Place the table anywhere on the chart:
Top Left
Top Right
Bottom Left
Bottom Right
• Manual Sector Selection
Choose the SPDR sector ETF used as the sector benchmark for relative performance analysis.
█ DATA INTERPRETATION
Positive values indicate that the asset is outperforming its benchmark or sector.
Negative values indicate underperformance, signaling weaker relative strength compared to the benchmark or sector being analyzed.
█ TIMEFRAME NOTES
Since the indicator analyzes historical performance up to 180 calendar days, the following timeframes are recommended:
1D
4H
1H
On very low timeframes (such as 1-minute or 5-minute charts), TradingView may not have enough historical bars loaded to calculate the longest lookback periods. In these cases, some cells may display "n/a" (data not available).
Ideal for quickly identifying assets that are demonstrating superior relative strength versus both the broader market and their respective sector, regardless of the chart timeframe being used.
指标
指标
指标
TrendPulse: 9 EMA + VWAP Continuation StrategyTrendPulse combines 9 EMA trend structure, VWAP positioning, and continuation logic into a chart-ready intraday strategy. It includes fully customizable visual aids, signal markers, dashboard metrics, and active trade overlays for entry, stop, target, and break-even visualization, allowing traders to tailor the display to their own strategy and charting preferences. Trade overlays appear only while a position is active and automatically disappear once the trade is closed to help keep charts clean and uncluttered.
TrendPulse combines 9 EMA trend structure, VWAP positioning, and continuation logic into a chart-ready intraday strategy built for traders who want both structure and flexibility. The script is designed to help identify trend alignment, continuation opportunities, and trade management levels while keeping the chart readable and customizable.
This strategy includes fully customizable visual aids, signal markers, dashboard metrics, and active trade overlays for entry, stop, target, and break-even visualization. All visual elements can be adjusted to better fit individual trading styles, chart layouts, and preferred market conditions. Active trade overlays appear only while a position is live and automatically disappear after the trade closes, helping reduce chart clutter.
How It Works
TrendPulse evaluates price structure using a combination of:
fast and slow EMA trend bias
a dynamic basis line
9 EMA context
VWAP positioning
volatility and regime conditions
volume and liquidity filters
optional benchmark confirmation
The strategy looks for breakout and continuation behavior when market conditions support trend movement. It is especially useful for traders who want a structured framework for momentum continuation setups while still having the ability to customize the chart presentation.
Key Features
9 EMA and VWAP overlays
Dynamic basis line with bullish, bearish, and neutral coloring
Breakout and continuation entry logic
Relative volume and dollar-volume liquidity filters
Optional market benchmark filter
Risk-based position sizing
Stop, target, and break-even trade overlays
Optional ATR-based trailing logic
Compact on-chart dashboard
Tiny buy/sell triangle signal markers
Fully customizable colors and visibility controls
Trade overlays shown only during active positions to reduce clutter
How To Use It
TrendPulse is best used as a structured intraday tool for identifying trend continuation conditions.
General long idea Look for:
bullish trend alignment
price holding above important structure
supportive VWAP positioning
improving momentum or continuation behavior
sufficient liquidity and relative volume
General short idea Look for:
bearish trend alignment
price staying below important structure
weak VWAP positioning
downside continuation behavior
sufficient liquidity and relative volume
Entry logic
Depending on the selected mode, the strategy can look for:
Breakout setups
Continuation setups
or Both
This lets traders adapt the script to different market conditions and personal preferences.
Risk management
The script can calculate:
entry
stop
target
optional break-even behavior
optional trailing behavior
The active trade lines are shown only while a trade is open, then removed automatically once the position closes so the chart stays clean.
Dashboard Guide
The TrendPulse dashboard is designed to help traders quickly assess market condition, directional quality, momentum participation, and whether price may already be extended.
State
Shows the current market regime: Trend, Expand, Quiet, or Noisy.
Trend suggests cleaner directional structure and better follow-through potential.
Expand suggests volatility is increasing, which can support strong momentum but also faster reversals.
Quiet suggests a slower or compressed market where breakouts may fail more often.
Noisy suggests mixed or unstable behavior with less reliable continuation.
How to use it: This is a context filter. Trend and Expand conditions are generally more favorable than Quiet or Noisy conditions for momentum-based setups.
L / S
Shows the current Long Score and Short Score.
A higher Long Score suggests stronger bullish alignment.
A higher Short Score suggests stronger bearish alignment.
If both are weak or close together, directional edge may be limited.
How to use it: Helps identify which side has better structure and quality. A clear score imbalance can support directional bias, while similar or weak scores may suggest patience.
Chop
Measures how choppy or directional recent price action has been.
Lower values generally suggest cleaner trend behavior.
Higher values usually suggest more back-and-forth movement.
How to use it: Lower Chop is generally more favorable for continuation-style setups. Higher Chop often means a greater chance of fakeouts, failed breakouts, or messy follow-through.
ATR Exp
Shows the ATR expansion ratio, which compares current volatility to its recent average.
Higher values suggest the market is becoming more active.
Lower values suggest a quieter or less energized environment.
How to use it: Helps gauge whether a move has enough energy behind it to continue. Rising ATR Exp can support momentum, but very high expansion can also mean the move is becoming aggressive and may be approaching exhaustion.
Trend
Shows the current directional bias: Bull, Bear, or Flat.
Bull favors long-side continuation thinking.
Bear favors short-side continuation thinking.
Flat suggests weaker directional edge.
How to use it: A quick directional filter to help traders stay aligned with broader short-term structure instead of trading against it.
% vs 9 EMA
Shows the percent distance of price from the 9 EMA.
Positive values mean price is above the 9 EMA.
Negative values mean price is below the 9 EMA.
How to use it: Helps judge short-term extension. The farther price moves from the 9 EMA, the more likely it may be becoming overextended or moving toward a capitulation/exhaustion phase. Smaller distances often reflect healthier continuation positioning, while larger distances can warn against chasing entries too late.
% vs VWAP
Shows the percent distance of price from VWAP.
Positive values mean price is above VWAP.
Negative values mean price is below VWAP.
How to use it: Helps judge how far price has moved from the session’s average traded value. A modest distance can support trend control, but a large distance may indicate emotional expansion, late-stage extension, or growing mean-reversion risk. If both the 9 EMA and VWAP distances are highly stretched in the same direction, the move may be strong but also increasingly vulnerable to pullback or exhaustion.
Vol
Shows relative volume compared with recent average volume.
Higher values suggest stronger participation.
Lower values suggest weaker participation.
How to use it: Stronger volume can support the credibility of a breakout or continuation move. Weak volume may mean the setup looks good visually but lacks enough participation to sustain follow-through.
Best practice
The dashboard works best when read as a group rather than field by field. For example, Trend or Expand state, strong directional score imbalance, lower Chop, healthy ATR expansion, and supportive volume can point to a cleaner continuation environment. On the other hand, high Chop, weak score separation, and very stretched distance from both the 9 EMA and VWAP may suggest caution, delayed entries, or increased exhaustion risk.
Customization
One of the main goals of TrendPulse is flexibility.
Users can customize:
visibility of moving averages, basis, channels, VWAP, and 9 EMA
signal marker colors
dashboard location, size, and theme
trade overlay colors
score label visibility and sizing
background regime highlights
This allows traders to simplify the chart or make it more information-rich depending on how they use it.
Best Instruments
TrendPulse is generally best suited for:
liquid stocks
active ETFs
high-volume intraday names
momentum-focused equities
It tends to be more useful on instruments where:
volume is meaningful
spreads are not excessive
VWAP and momentum behavior matter intraday
It may be less effective on:
illiquid symbols
very low-volume names
assets with inconsistent intraday movement
heavily erratic charts with poor liquidity
Best Timeframes
This strategy is primarily designed for intraday use.
Most suitable timeframes:
1 minute
3 minute
5 minute
15 minute
Some traders may also experiment with:
30 minute
As always, settings may need adjustment depending on the instrument and timeframe being traded.
Notes
This script is intended as a structured strategy and visualization tool.
Users should test settings across different symbols and timeframes.
No single parameter set is ideal for every market.
Traders may prefer different combinations of breakout, continuation, and filtering logic depending on their own process.
Disclaimer
For educational and research purposes only.
This script does not constitute financial advice.
Always forward test, validate settings, and manage risk appropriately before using any strategy in live markets.
If you find TrendPulse useful, consider saving it, sharing feedback, and adapting the visuals and filters to match your own trading workflow.
策略
Ichimoku Cloud Thickness SpectrumOverview
Ichimoku Cloud Thickness Spectrum is an overlay for studying the geometry and relative thickness of the projected Ichimoku cloud. It plots the classical Tenkan-sen, Kijun-sen, Senkou Span A, Senkou Span B, and an optional Chikou Span, then adds a configurable research layer derived only from the distance and relationship between the two Senkou spans.
The script is designed to make three questions easier to inspect:
- How wide is the projected cloud in a unit suited to the current instrument?
- How unusual is that width compared with the cloud's own recent history?
- Is the smoothed cloud thickness expanding, contracting, or comparatively stable?
This is a context and visualization tool. It does not generate buy or sell instructions, forecasts, targets, stops, position sizing, or performance claims.
Original contribution
Cloud thickness is a known Ichimoku concept. The original contribution of this implementation is the integrated workflow used to transform raw Span A/B distance into a configurable research display:
1. Selectable ATR, percentage, or raw-price normalization.
2. Rolling empirical percentile ranking of the selected thickness measure.
3. Percentile-weighted cloud opacity and an optional dense inner core.
4. A stabilized expansion/contraction spine with a near-zero denominator floor and release-ratio hysteresis.
5. Confirmed thin-cloud seams filtered by thin-state membership, two-sided prominence, and minimum spacing.
6. Directional projected-cloud twist markers.
7. Separate handling of the newly calculated forward cloud and the historically shifted cloud aligned with the current price bar.
8. A bilingual confirmed/live readout, confirmed-bar alerts, and Data Window diagnostics.
The classical Ichimoku formulas are standard, widely known calculations. The percentile, motion-state, seam-filtering, visualization, readout, and alert logic were implemented for this script. No source code from another Community Script is reused.
Core calculations
- Tenkan-sen: midpoint of the highest high and lowest low over the Conversion length.
- Kijun-sen: midpoint of the highest high and lowest low over the Base length.
- Senkou Span A: average of Tenkan-sen and Kijun-sen.
- Senkou Span B: midpoint of the highest high and lowest low over the Span B length.
- Raw cloud thickness: absolute distance between Senkou Span A and Senkou Span B.
- Cloud midpoint: average of Senkou Span A and Senkou Span B.
Thickness modes and percentile ranking
- ATR: raw cloud thickness divided by ATR. This is the default and relates cloud width to recent price range.
- Percent: raw cloud thickness divided by the absolute close, multiplied by 100.
- Raw: the unnormalized price distance between Span A and Span B.
The selected measure is ranked against its own rolling history with a percent-rank calculation. A reading near 10 is relatively thin within the selected lookback, while a reading near 90 is relatively thick. The result is specific to the symbol, timeframe, settings, and available history; it is not a universal market-strength score.
Stabilized expansion and contraction
The selected thickness measure is smoothed with an EMA. A one-bar percentage change is then calculated. Near a cloud twist, Span A and Span B can converge and the previous thickness can approach zero, which can make an unrestricted percentage change unstable. To reduce this effect, the denominator cannot fall below a configurable fraction of a slower EMA baseline.
The state classifier also uses release-ratio hysteresis. Entry into expansion or contraction requires the main sensitivity threshold, while release from an existing state uses a smaller configurable threshold. This reduces bar-to-bar color chatter. The state uses the previous bar and does not rely on intrabar-only persistent memory.
Visual interpretation
- Teal cloud: Senkou Span A is at or above Senkou Span B.
- Red cloud: Senkou Span A is below Senkou Span B.
- More transparent cloud: lower thickness percentile.
- More visible cloud: higher thickness percentile.
- Dense inner core: optional emphasis when thickness percentile is above the selected threshold.
- Blue midpoint spine: thickness is in an expanding state.
- Purple midpoint spine: thickness is in a contracting state.
- Gray midpoint spine: thickness motion is inside the release threshold.
- Yellow upward/downward triangle: confirmed bullish/bearish projected-cloud twist.
- Small pink X: confirmed local thin-cloud seam that passed the thin-state, prominence, and spacing filters.
The cloud opacity and inner core encode relative thickness only. They do not guarantee support, resistance, continuation, reversal, or breakout behavior.
Filtered thin-cloud seams
A thin seam is not every low-thickness bar. It must be a confirmed local minimum in the thickness percentile, remain at or below the user-defined thin threshold, rebound by at least the selected prominence on both sides, and pass the minimum-spacing filter.
The pivot becomes known only after the selected number of right-side bars has closed. By default, a marker is not placed to the left of the bar where confirmation became available. The optional Allow seam placement before confirmation setting can place the marker at the original pivot-projected cloud position after confirmation. This changes the visual location only and does not make the event available earlier.
Projected cloud versus current aligned cloud
The newly calculated Span A and Span B values are displayed forward by the selected displacement. This positive offset shifts calculations made from current and historical bars into future chart positions; it does not access future market data.
Price-context calculations use a different alignment. They reference the historically shifted spans that are actually visible at the current price bar. Optional inside-cloud highlighting and current-cloud transition alerts therefore do not compare current price with the newly calculated cloud plotted in the future.
Forward readout, alerts, and research outputs
The three-line readout displays the thickness state and percentile, the selected normalized thickness value and motion state, and the future-cloud direction with Live or Confirmed status. English is the default. Japanese changes the readout, hover tooltip, and combined dynamic alert messages; input names and tooltips are bilingual.
Use confirmed values in readout is enabled by default. While the latest realtime bar is open, the label uses the previous closed bar. Once the latest bar is confirmed, it uses that bar. Users can disable the setting to inspect the open realtime bar, in which case the label shows Live and may change until close.
Confirmed-bar alert conditions are available for projected twists, thin/thick state transitions, current-cloud position transitions, expansion/contraction changes, and filtered thin seams. A combined bilingual alert is also available through Any alert() function call. Additional raw, normalized, percentile, motion, seam, current-cloud, direction, and state values are provided in the Data Window.
Realtime behavior and limitations
The script does not use request.security(), lookahead, or future bars. On an open realtime bar, Tenkan-sen, Kijun-sen, the projected spans, cloud opacity, and motion state can update as the bar's high, low, and close change. Confirmed historical values do not use intrabar-only memory. The supplied alert conditions are evaluated on confirmed bars, and the combined alert() call uses once-per-bar-close frequency.
Percentile readings require sufficient history and can show a warming-up state. Results depend on the symbol, timeframe, Ichimoku lengths, normalization basis, percentile lookback, and available data. Non-time-based charts and sparse datasets may behave differently from standard time-based candles.
A thick cloud is not automatically strong support or resistance. A thin or contracting cloud is not a breakout prediction. This script provides research context and should be combined with the user's own analysis and risk process.
日本語概要
Ichimoku Cloud Thickness Spectrumは、一目均衡表の将来雲について、方向だけでなく、厚み、履歴内の相対順位、拡大・収縮状態を研究するためのオーバーレイ型インジケーターです。転換線、基準線、先行スパン1・2、任意の遅行スパンを表示し、追加分析はすべて先行スパン1と2の距離および関係から算出します。
主な特徴
- 雲の厚みをATR比率、価格比率、価格差から選択可能。
- 選択した厚みを直近履歴内のパーセンタイルへ変換。
- 厚み順位を雲の透明度と任意の高密度インナーコアへ反映。
- ゼロ近傍の分母下限とヒステリシスを備えた拡大・収縮中心線。
- 薄い状態、左右プロミネンス、最小間隔を満たした確認済み局所薄化点。
- 上向き・下向きを分けた確認済み将来雲ツイスト。
- 将来へ描画する新規計算雲と、現在価格バーへ整列した過去計算雲を分離。
- 英語・日本語を切り替えられる将来ラベル、確定足アラート、データウィンドウ研究値。
雲の厚みという概念自体は既知です。本実装の差異は、生の雲幅を正規化し、履歴順位化し、厚みの動きを安定化し、薄化点を確認・選別し、それらを一貫した視覚表現と研究出力へ統合している点です。標準一目均衡表の計算式は広く知られた古典計算であり、追加した順位化、状態判定、薄化点フィルター、表示、ラベル、アラートのロジックは本スクリプト用に実装しています。他のCommunity Scriptのソースコードは再利用していません。
表示の読み方
- 青緑の雲: 先行スパン1が先行スパン2以上。
- 赤の雲: 先行スパン1が先行スパン2未満。
- 透明度が高い雲: 厚み順位が低い。
- 濃い雲とインナーコア: 厚み順位が高い。
- 青の中心線: 厚みが拡大状態。
- 紫の中心線: 厚みが収縮状態。
- 灰色の中心線: 安定状態。
- 黄色の三角: 確認済みの上向き・下向き将来雲ツイスト。
- 小さなピンクのX: 条件を満たした確認済み局所薄化点。
厚み順位は銘柄、時間足、設定、参照期間、利用可能な履歴に依存する相対値です。雲の濃さやインナーコアは支持抵抗の保証ではありません。
確認、シフト、ライブ動作
将来雲は、現在および過去バーから計算した先行スパンを正のoffsetで将来位置へ描画したもので、未来データを取得していません。現在価格との関係には、現在バー上に実際に整列している過去計算の先行スパンを使用します。
薄化点は右側に指定本数の確定バーを必要とするため、認識に意図的な遅延があります。初期設定では確認可能になったバーより左側へ表示しません。任意設定で元の投影位置へ表示する場合も、認識自体は確認後です。
未確定リアルタイムバーでは、高値・安値・終値の更新に伴い、転換線、基準線、先行スパン、雲の透明度、拡大・収縮状態が終値確定まで変化する場合があります。将来ラベルは初期設定で確定値を使用し、用意したアラートは確定足で判定します。
本インジケーターは研究・可視化ツールです。売買指示、将来予測、目標価格、損切り、ポジションサイズ、成績統計、投資助言を提供しません。厚い雲を必ず強い支持抵抗、薄い雲や収縮中の雲を必ずブレイクする状態とは定義していません。 指标
指标
指标
指标
Investory MTF Dashboard [ULTIMATE]//@version=6
indicator('Investory MTF Dashboard ', overlay = true)
// ==========================================
// --- PARAMETER INPUT ---
// ==========================================
rsi_len = input.int(14, 'RSI Length', group = 'Indikator')
ma_len = input.int(14, 'MA-RSI Length', group = 'Indikator')
fast_len = input.int(12, 'MACD Fast Length', group = 'Indikator')
slow_len = input.int(26, 'MACD Slow Length', group = 'Indikator')
sig_len = input.int(9, 'MACD Signal Smoothing', group = 'Indikator')
pos_db = input.string('Top Right', 'Posisi Dashboard', group = 'Tampilan', options = )
f_size = input.string('Normal', 'Ukuran Font', group = 'Tampilan', options = )
// --- FUNGSI POSISI & SIZE ---
f_get_pos(p_txt) =>
switch p_txt
'Top Right' => position.top_right
'Top Left' => position.top_left
'Bottom Right' => position.bottom_right
'Bottom Left' => position.bottom_left
=> position.top_right
ts = f_size == 'Tiny' ? size.tiny : f_size == 'Small' ? size.small : f_size == 'Large' ? size.large : size.normal
// ==========================================
// --- FUNGSI HITUNG (CORE) ---
// ==========================================
f_calc_indicators() =>
// RSI
rsi_v = ta.rsi(close, rsi_len)
ma_v = ta.sma(rsi_v, ma_len)
// MACD
= ta.macd(close, fast_len, slow_len, sig_len)
[rsi_v, rsi_v , ma_v, ma_v , hi, hi , ml]
// ==========================================
// --- REQUEST DATA DENGAN OPTIMASI ---
// ==========================================
// Menggunakan tuple untuk mengambil semua data sekaligus per timeframe
= request.security(syminfo.tickerid, '5', f_calc_indicators(), gaps = barmerge.gaps_off)
= request.security(syminfo.tickerid, '15', f_calc_indicators(), gaps = barmerge.gaps_off)
= request.security(syminfo.tickerid, '60', f_calc_indicators(), gaps = barmerge.gaps_off)
= request.security(syminfo.tickerid, '240', f_calc_indicators(), gaps = barmerge.gaps_off)
= request.security(syminfo.tickerid, 'D', f_calc_indicators(), gaps = barmerge.gaps_off, lookahead = barmerge.lookahead_on)
// ==========================================
// --- RENDER DASHBOARD ---
// ==========================================
// Kolom dikurangi menjadi 9 (0 sampai 8) karena kolom A/D dihapus
var table main_db = table.new(f_get_pos(pos_db), 9, 7, frame_color = color.new(color.white, 80), frame_width = 1, border_color = color.new(color.white, 80), border_width = 1)
if barstate.islast
// Warna tema
bg_main = color.new(color.black, 10)
h_bg = color.new(color.gray, 20)
l_bg = color.new(color.gray, 50)
// Header
table.cell(main_db, 0, 0, 'TF', bgcolor = h_bg, text_color = color.white, text_size = ts)
table.cell(main_db, 1, 0, 'RSI ANALYSIS', bgcolor = h_bg, text_color = color.white, text_size = ts)
table.merge_cells(main_db, 1, 0, 4, 0)
table.cell(main_db, 5, 0, 'MACD TREND', bgcolor = h_bg, text_color = color.white, text_size = ts)
table.merge_cells(main_db, 5, 0, 8, 0)
// Sub-Header
labels = array.from('-', 'RSI', 'MA', 'DIFF', 'PWR', 'HIST', 'SIG', 'PWR', 'TREND')
for j = 0 to 8 by 1
table.cell(main_db, j, 1, array.get(labels, j), bgcolor = l_bg, text_color = color.white, text_size = ts)
// Data Processing
tfs = array.from('5M', '15M', '1H', '4H', 'D')
rsis = array.from(r5, r15, r60, r240, rD)
rsips = array.from(rp5, rp15, rp60, rp240, rpD)
mas = array.from(m5, m15, m60, m240, mD)
maps = array.from(mp5, mp15, mp60, mp240, mpD)
hists = array.from(h5, h15, h60, h240, hD)
hps = array.from(hp5, hp15, hp60, hp240, hpD)
ml_s = array.from(ml5, ml15, ml60, ml240, mlD)
for i = 0 to 4 by 1
row = i + 2
// Ambil data dan handle nilai NaN (Data belum load)
rv = array.get(rsis, i)
rpv = array.get(rsips, i)
mv = array.get(mas, i)
mpv = array.get(maps, i)
hv = array.get(hists, i)
hpv = array.get(hps, i)
mlv = array.get(ml_s, i)
if not na(rv) // Cek jika data tersedia
color rc = rv > rpv ? color.lime : color.red
color mc = mv > mpv ? color.lime : color.red
table.cell(main_db, 0, row, array.get(tfs, i), bgcolor = bg_main, text_color = color.white, text_size = ts)
table.cell(main_db, 1, row, str.tostring(rv, '#.#'), bgcolor = bg_main, text_color = rc, text_size = ts)
table.cell(main_db, 2, row, str.tostring(mv, '#.#'), bgcolor = bg_main, text_color = mc, text_size = ts)
diff = rv - mv
table.cell(main_db, 3, row, str.tostring(diff, '#.#'), bgcolor = bg_main, text_color = diff >= 0 ? color.lime : color.red, text_size = ts)
table.cell(main_db, 4, row, diff >= 0 ? '▲' : '▼', bgcolor = bg_main, text_color = diff >= 0 ? color.lime : color.red, text_size = ts)
// MACD Logic
table.cell(main_db, 5, row, str.tostring(hv, '#.##'), bgcolor = bg_main, text_color = hv >= 0 ? color.lime : color.red, text_size = ts)
table.cell(main_db, 6, row, mlv > 0 ? '⬆' : '⬇', bgcolor = bg_main, text_color = mlv > 0 ? color.lime : color.red, text_size = ts)
table.cell(main_db, 7, row, hv > hpv ? '▲' : '▼', bgcolor = bg_main, text_color = hv > hpv ? color.lime : color.red, text_size = ts)
// TREND Text
string txtT = mlv > 0 and hv > hpv ? 'Bull' : mlv < 0 and hv < hpv ? 'Bear' : mlv > 0 ? 'COR' : 'REV'
color colT = txtT == 'Bull' ? color.lime : txtT == 'Bear' ? color.red : txtT == 'COR' ? color.yellow : color.orange
table.cell(main_db, 8, row, txtT, bgcolor = bg_main, text_color = colT, text_size = ts)
else // Tampilan jika data sedang loading
table.cell(main_db, 0, row, array.get(tfs, i), bgcolor = bg_main, text_color = color.white, text_size = ts)
table.cell(main_db, 1, row, 'Wait...', bgcolor = bg_main, text_color = color.gray, text_size = ts)
table.merge_cells(main_db, 1, row, 8, row)
指标
Breakout & Retest Entry Signals & the Break-vs-RetestOVERVIEW
"Wait for the retest" is the most repeated piece of advice in breakout trading. It is also, as far as I can tell, completely untested by the people who repeat it.
This tool settles it — on your instrument, with your settings — by taking BOTH entries on the SAME breaks and grading them against the same control:
PER-TRADE EDGE expectancy vs control · n
Break entry +0.09R +0.02R · n=412
Retest — CLEAN +0.31R +0.02R · n=118
Retest — DEEP -0.04R +0.02R · n=76
clean vs deep (t) 3.41 CLEAN IS BETTER
BUT — HOW OFTEN DO YOU GET IT?
Breaks that ever retested 47.0% (194/412)
after a run of >3 closes 31.2% vs 55.8% otherwise
False-break rate 13.3%
EXPECTED VALUE PER BREAK
Take every break +0.09R (100% of breaks)
Wait for the retest +0.06R (47% of breaks)
VERDICT NO DIFFERENCE — pick either
That last block is the entire point. A better per-trade edge is worthless if you only get the trade half the time — so the WAIT policy is scored as P(retest) x E , because on every break that never retests you get NOTHING. The two are then compared with a significance test, and the verdict is allowed to be "no difference".
It is a research and framing tool. NOT a strategy, NOT a signal service, NOT a validated edge.
WHAT IT ALREADY FOUND — measured live on NIFTY futures
timeframe breaks retested false breaks verdict
1m 334 49.4% 29.6% NO DIFFERENCE
3m 323 53.3% 13.6% NO DIFFERENCE
5m 328 53.4% 14.0% NO DIFFERENCE
15m 306 51.0% 12.7% TAKE THE BREAK
1h 274 50.0% 16.4% NO DIFFERENCE
TWO THINGS JUMP OUT.
The retest rate is 50-53% on every timeframe. Bulkowski, measuring throwbacks across 10,348 chart patterns on US daily stocks, found 50-60%. A completely different market, a completely different method, and the same number. That is a real phenomenon, not an artefact of the detector.
The false-break rate is 26.5% on the 1m and 13-15% everywhere else. THE ONE-MINUTE BREAK IS TWICE AS LIKELY TO BE A LIE. That is not folklore, it is this instrument's own number, and it is exactly the kind of thing a trader should know before choosing a timeframe.
And the verdict, on four of the five: NO DIFFERENCE — pick either. Once the geometry is honest and the test is a real one, the great break-versus-retest argument simply does not resolve on this instrument at most speeds. On the 15m it does resolve — and it says TAKE THE BREAK, which is the opposite of what almost everyone will tell you.
That is what a measurement looks like. It disagrees with the folklore on one timeframe, refuses to take a side on four others, and does not care what you were hoping for. No tool that needs to sell you a signal would ever print "NO DIFFERENCE".
THE ONE THING EVERYONE GETS BACKWARDS
Thomas Bulkowski measured throwbacks and pullbacks across 10,348 chart patterns. His finding:
"Do throwbacks hurt performance? YES: 97% of the time chart patterns with upward breakouts
perform better post-breakout WITHOUT a throwback."
"Do pullbacks hurt performance? YES: 91% of chart pattern types with downward breakouts
perform better if a pullback does NOT occur."
Read that again. The retest is not a gift. It is a SYMPTOM — evidence that the move is weak, that supply came back, that the break did not have the strength to run.
And yet "wait for the retest" is good advice for a completely different reason: it gives you a better price and a tighter stop.
BOTH ARE TRUE AT ONCE. They are two opposing effects on the same trade, and they have never been put on one scale and netted out. That is what this script does. The retest may still win — a better entry can outweigh a weaker move — but nobody has ever checked, and the answer is different on every instrument and every timeframe.
A CLEAN RETEST AND A DEEP ONE ARE NOT THE SAME EVENT
Bulkowski again, and this is his sharpest single finding on the subject: during a throwback, if price REMAINS ABOVE the breakout price the subsequent rise averages 40%. If it drops BELOW the breakout price and then recovers, the rise averages 29%. That is 400 samples versus 2,767.
Pooling those two throws away the strongest signal in the whole idea. So they are separated:
CLEAN retest — price came back and touched the level, but never CLOSED back through it.
DEEP retest — price CLOSED back through the level, then recovered.
They are graded separately, tested against each other, and labelled separately on the chart. If clean beats deep on your instrument, then "wait for the retest" is not one rule — it is two, and only one of them works.
AND CAN YOU SEE IT COMING?
The real, unpriced cost of a WAIT policy is that roughly half the time you never get filled. So it matters enormously whether you can predict which breaks will retest.
Bulkowski found that if price has more than three consecutively higher closes ending the day before the breakout, the throwback probability drops materially. So the panel reports the retest rate SPLIT BY THAT:
after a run of >3 closes 31.2% vs 55.8% otherwise
If the split is real on your instrument, then after a strong run into the break you should simply TAKE IT — because the retest you are waiting for is probably never coming.
IS YOUR VOLUME FILTER EARNING ITS KEEP?
Every trader is taught that a breakout must be confirmed by volume. Bulkowski's volume study says that after an ABOVE-average-volume breakout, FAILURES DOUBLE and the likelihood of a throwback TRIPLES, while the move itself is barely better.
That is testable — but only if the low-volume breaks are allowed into the sample. So VOLUME IS NOT A GATE ON THE RECORD. Every break is recorded; volume gates only the SIGNAL. The panel then reports what your filter is actually worth:
Break ON volume +0.11R n=246
Break OFF volume +0.06R n=166
on vs off (t) 0.82 no difference — it is doing nothing
The record is a fact about the market. The filter is a decision about the trade. They are kept apart, and this is what happens when you stop assuming and start measuring.
IDENTICAL GEOMETRY — and why this is not a detail
The target used to be THE NEXT OPPOSING LEVEL. That quietly destroyed the entire experiment.
The BREAK entry sits PAST the level (it closed through it). The RETEST entry sits BACK AT the level. So the retest is systematically FARTHER from the next opposing level, and was therefore being handed a BIGGER R:R for the SAME RISK — on every single trade, by construction. Live, that produced an R:R of 5.0 on one timeframe and 0.66 on another, and the on-chart key was cheerfully claiming "identical geometry" while the geometry was tilted toward the retest.
The trade now uses a FIXED R multiple, identical for the break, the retest and the control. The next opposing level is still drawn, and still tested — separately, as a descriptive statistic, with its hit rate reported next to its distance in R.
THE ANTI-BIAS GUARDS
ENTRY IS THE CLOSE, for both entries and for the control. Entering the retest AT the level — a better price than the close — while the break enters at its close would hand the retest a free head start on every trade, and settle the oldest argument in trading by rigging it.
THE CONTROL IS DIRECTION-MATCHED. Breaks run with the trend, so a direction-skewed event set measured against a symmetric 50/50 control inherits the drift for free and calls it an edge. Longs are compared only with control longs, shorts only with control shorts, and the control is blended back using the events' OWN direction mix.
EVERY VERDICT IS A TEST, NOT A COMPARISON. Break-vs-wait, clean-vs-deep, volume-on-vs-off — each is a Welch t-test that has to clear |t| > 1.96 before it is allowed to be a finding. For the wait policy, the variance of P(retest) x E is propagated by the delta method, because it is a product of two estimates and both carry error. A verdict that flips on a tenth of an R is not a verdict, it is noise wearing a costume.
Both barriers on one bar: the STOP is assumed first — conservative, and the only assumption that cannot flatter the result. Unresolved trades at the horizon are marked to market, not booked as losses.
THE LEVELS
Levels come from the extrema of a KERNEL-SMOOTHED price series (Nadaraya-Watson) rather than raw pivots, so they track the structure rather than the noise. A break requires a CLOSE beyond the level with displacement, not a wick. A false break is one that closes back inside quickly. All of it is computed on confirmed bars; the kernel is causal and never looks forward.
NON-REPAINT
The kernel confirms an extremum a half-window late, so a level appears some bars AFTER the swing that created it. That lag is the price of not repainting and it is paid deliberately. Levels, breaks, false breaks, retests, signals and every calibration event are computed on CONFIRMED bars only. Nothing is drawn and then moved.
DATA AND SCOPE
Any symbol, any timeframe. ATR-normalised throughout. Volume improves the SIGNAL but is not required, and it never gates the RECORD.
EXPORTS (Data Window — consume from other scripts via input.source())
EXP_Level, EXP_Break, EXP_FalseBreak, EXP_Retest, EXP_Entry, EXP_Stop, EXP_Target, EXP_NextLevel, EXP_WaitEdge
CONCEPT CREDIT
Support/resistance, polarity and the breakout-retest idea are long-standing public trading concepts with no single author; the written tradition runs through Charles Dow, Richard Wyckoff and Edwards & Magee. The formal TRADING-RANGE BREAK was first tested at scale by William Brock, Josef Lakonishok and Blake LeBaron, Journal of Finance 47(5), 1992 — and their results were later shown to be vulnerable to data-snooping (Sullivan, Timmermann and White, 1999), which is exactly why this tool measures the rule on YOUR instrument rather than asserting it.
The throwback and pullback statistics that motivate the clean/deep split, the run-length predictor and the volume test are from Thomas Bulkowski ("Encyclopedia of Chart Patterns"; thepatternsite.com). His numbers are measured on US daily stocks. Whether they hold on YOUR instrument is precisely the question this script exists to answer — and it may well answer "no".
Nadaraya-Watson kernel regression — Nadaraya and Watson (1964); its use for technical pattern recognition — Lo, Mamaysky and Wang, Journal of Finance 55(4), 2000. Triple-barrier forward labelling — Marcos Lopez de Prado. Welch's t-test — B. L. Welch. ATR — J. Welles Wilder.
The break-vs-wait availability weighting, the delta-method significance test, the clean/deep retest split, the volume-filter test and the direction-matched control are the author's own. Clean-room implementation; no third-party Pine code is reused. Not affiliated with, nor endorsed by, any of the above.
HONESTY AND LIMITATIONS
Calibration is IN-SAMPLE, with no costs or slippage, and uses overlapping windows. A proven in-sample edge is NOT a guarantee out-of-sample. Real fills, spreads and commissions will reduce it — and they will hurt the break entry more than the retest entry, because the break enters into momentum.
Bulkowski's throwback statistics are measured on US daily stocks over decades. They are the reason the questions are asked. They are NOT the answer, and this tool will tell you so if your instrument disagrees.
The verdict is allowed to be "NO DIFFERENCE — pick either", and on many instruments it will be. That is a real result. A tool that cannot report its own failure is an advertisement, not a measurement.
Nothing in this script predicts price.
DISCLAIMER
Research and educational tool only. NOT financial advice, NOT a recommendation, and NO guarantee of results. Entry, stop and target output is arithmetic, not advice. Trading carries risk of loss. Test out-of-sample and make your own decisions. The author accepts no liability for any use.
指标
指标
Anchored VWAP Hand-Off 2.03Anchored VWAP Hand-Off 2.03 is a multi-leg cascading VWAP indicator. It starts with a classic Anchored VWAP from a user-selected date (Leg 1), then automatically creates new "hand-off" VWAP legs (up to 24) every time price crosses the previous VWAP. Each new leg resets its calculation from the bar of the cross, creating a dynamic chain of VWAPs that follow price action.
Key Features:
Source: Typically close (customizable).
Trigger: EMA(1) of the source (essentially price itself).
Color Logic: Each VWAP leg changes color based on whether price is above (bullish) or below (bearish) it. Supports unified color mode.
Up to 24 cascading legs (user controls how many are shown).
Visual anchor line at the start date.
Main Use Cases:
Dynamic Support & Resistance
The multiple VWAP legs act as adaptive moving support/resistance zones. Traders watch for bounces off these lines or breaks through several legs at once.
Trend Strength & Momentum
Strong trends often break multiple hand-off legs quickly.
Choppy/consolidating markets produce many frequent hand-offs (more lines clustering).
Mean Reversion Setups
Price tends to return to the nearest active VWAP legs. Useful for fade-the-extreme strategies.
Breakout / Breakdown Confirmation
When price decisively breaks through several VWAP legs in sequence, it often signals a high-probability directional move.
Intraday & Swing Trading
Particularly popular on stocks, futures, and crypto for finding "fair value" shifts after news events or strong moves. The hand-off mechanism makes it more responsive than a single static anchored VWAP.
Visual Market Structure
Helps traders see how "value" is migrating over time as new legs form.
Best Timeframes: Works on all, but especially useful on 5min to daily charts.Pro Tip: Start with 8–12 legs. Too many legs can make the chart noisy. Use the anchor date on significant events (earnings, breakout days, macro events, etc.).This indicator is an evolution of standard Anchored VWAP, designed to solve the problem of a single anchored VWAP becoming stale after big moves.
指标
Institutional Sniper Signal (Clean v7)This indicator is a Multi-Timeframe (MTF) quantitative system designed to trade pullbacks and breakouts in alignment with the higher timeframe macro trend.
Rather than relying on a single indicator to dictate the trend, this script utilizes a Consensus Voting Mashup. By aggregating data from 15 different structural and momentum indicators on a higher timeframe, it calculates a definitive directional bias, filtering out the noise of ranging markets.
⚙️ CORE MECHANICS (The Consensus Logic):
To avoid repainting and false signals, the script operates in three strict phases:
The Macro Consensus (4H Closed Candle): The engine pulls data from the last closed 4H candle across 15 distinct tools (including EMA 200, EMA 50, WMA 100, Parabolic SAR, Ichimoku Tenkan/Kijun, DEMA, TEMA, and standard deviation bands). Each indicator gets 1 "vote" (Buy or Sell). A trade is only authorized if a strict user-defined threshold is met (e.g., 10 out of 15 indicators agreeing on the same direction).
The Pullback Trigger (1H): Once the macro consensus is formed, the script monitors the 1H timeframe. It waits for the local price to retrace and touch the 1H 20 EMA, identifying a potential discounted entry area.
The Fractal Breakout (Execution): It does not enter at market price. Instead, it calculates a 5-bar Fractal High (for longs) or Fractal Low (for shorts). It then plots visual pending order lines (Buy Stop / Sell Stop) at these fractal extremes, adding a customizable price offset to avoid fakeouts.
🛡️ RISK MANAGEMENT:
ATR Stop Loss: The stop loss is dynamically plotted at the opposite fractal, buffered by an ATR multiplier to allow the trade to breathe.
Daily Limit: Includes a daily signal limiter to prevent overtrading during high-volatility sideways days.
🛠️ HOW TO USE:
Apply the script to your execution timeframe (e.g., 15m or 30m). Adjust the "Offset" input based on your asset (e.g., 0.0002 for Forex pairs or 2.0 for US Indices). When a valid consensus is met and the price pulls back, the indicator will plot horizontal lines with precise labels for your pending Entry and Stop Loss.
🇧🇷 (PORTUGUÊS)
Este indicador é um sistema quantitativo Multi-Timeframe (MTF) projetado para operar pullbacks e rompimentos alinhados com a tendência macro.
Em vez de usar apenas um indicador para definir a tendência, este script utiliza um Sistema de Votação por Consenso. Ele agrega dados de 15 indicadores diferentes no tempo gráfico maior para calcular um viés direcional definitivo.
⚙️ COMO FUNCIONA A LÓGICA:
Para evitar repintura (repainting), o script opera em 3 fases:
O Consenso Macro (Vela Fechada H4): O motor puxa os dados da última vela fechada de H4 em 15 ferramentas diferentes (EMA 200, Ichimoku, DEMA, SAR, etc). Cada indicador dá 1 "voto". O setup só é armado se a maioria esmagadora (ex: 10 de 15) concordar com a direção.
O Gatilho de Pullback (H1): Com o consenso formado, o script espera o preço retrair e tocar na EMA 20 do H1 (área de desconto).
O Rompimento do Fractal (Entrada): O indicador não entra a mercado. Ele mapeia a Máxima ou Mínima dos últimos 5 candles (Fractal) e desenha uma linha de ordem pendente (Buy Stop / Sell Stop) com um recuo (offset) customizável.
COMO USAR: Adicione no seu gráfico de execução (ex: M15). Ajuste o valor do "Offset" nas configurações de acordo com o ativo (ex: digite 0.0002 se for Forex ou 2.0 se for Índice). Aguarde as linhas de entrada e Stop Loss aparecerem na tela para posicionar suas ordens pendentes. 指标
指标
指标
Machine Learning Price Bands Kernel Regression SignalsOVERVIEW
Every "AI band" on this platform draws two lines and asserts them. None of them can tell you how often price actually stays inside.
This one can — because it is built on a method that comes with a MATHEMATICAL COVERAGE GUARANTEE, and then it CHECKS WHETHER IT KEPT THE PROMISE, live, on your chart:
Coverage (empirical vs nominal) 89.1% vs 90% n = 20,266
Is the miss REAL? -0.9 pp z = -4.3 (real)
Verdict undercovering — real, but small
That is not a band. That is a prediction interval that has been audited, and it is the whole reason this tool exists.
It is a research and framing tool. NOT a strategy, NOT a signal service, NOT a validated edge.
THE MACHINE LEARNING, SPELLED OUT — no buzzwords, here is the actual model
1. NADARAYA-WATSON KERNEL REGRESSION. Non-parametric: no functional form is assumed, the data chooses the shape. Each past bar votes on the current estimate with a Gaussian weight that decays with distance. This is the same estimator Lo, Mamaysky and Wang used in the Journal of Finance to make chart-pattern recognition objective. It is real machine learning, and it is sixty years old.
The kernel here is CAUSAL. It only ever looks backwards. A centred kernel — the kind most "Nadaraya-Watson envelope" scripts use — peeks at bars that have not happened yet, and that is why their historical fit looks so much better than their live one.
2. BANDWIDTH BY PREDICTIVE MODEL SELECTION. The bandwidth h is the only real parameter, and it is not a magic number: several candidates are run in parallel and scored on their ROLLING ONE-STEP-AHEAD SQUARED ERROR. The winner is used. That is honest model selection — the criterion you would use to choose any forecaster — rather than a knob you turn until the chart looks nice.
3. CONFORMAL PREDICTION INTERVALS. The half-width is the (1-alpha) empirical quantile of the recent ABSOLUTE one-step-ahead errors. Under exchangeability this carries a FINITE-SAMPLE coverage guarantee, with NO distributional assumption at all: no normality, no GARCH, no volatility model. The model's own recent mistakes size the band — which is why it widens when the model starts being WRONG, not merely when price starts moving.
4. ADAPTIVE CONFORMAL INFERENCE — Gibbs and Candes, NeurIPS 2021.
Here is the problem with plain conformal prediction on markets, stated plainly: its guarantee holds under EXCHANGEABILITY, and financial returns are the textbook counterexample. Volatility CLUSTERS. So a residual quantile computed over a trailing window is always a step behind, the band is too narrow exactly when it matters, and the misses bunch together. Coverage lands quietly under nominal. Measured live on NIFTY futures before this was added: 89.1% against a nominal 90%, on the 1m, the 3m and the 1h, every one of them roughly four standard errors below target. Not a bug. The assumption breaking.
ACI makes the miscoverage level a LEARNED parameter:
alpha(t+1) = alpha(t) + gamma * (alpha - err(t))
Miss the interval and alpha falls, so the quantile rises and the band WIDENS. Cover it and alpha creeps back, so the band TIGHTENS. Long-run coverage provably converges to the target IRRESPECTIVE OF THE DATA GENERATING PROCESS — no exchangeability assumption anywhere.
A band that notices it is undercovering and fixes itself. Watch the alpha row: where it settles BELOW nominal is a direct measurement of how badly exchangeability fails on your instrument.
MEASURED, ON THE SAME INSTRUMENT, BEFORE AND AFTER:
timeframe plain conformal with ACI
1m 89.1% 90.1%
3m 89.1% 90.1%
5m 89.1% 90.1%
15m 90.1% 90.1%
1h 89.1% 90.1%
(nominal 90%)
Five timeframes, a four-standard-error undercoverage on four of them, closed. The binomial test now returns "calibrated — within sampling noise" and means it. That is not a backtest of a trading rule. That is a mathematical promise being kept, and being checked.
5. NORMALISED NONCONFORMITY — Papadopoulos et al. (2008), Lei et al. (2018).
The plain score |price - fit| is a SCALAR, which means the band is THE SAME WIDTH in a dead tape and in a crash. It therefore OVERCOVERS in calm and UNDERCOVERS in chaos — and the single marginal coverage figure is the average of those two errors, looking correct while being wrong in both directions.
Normalising divides each residual by a local scale estimate before taking the quantile, and multiplies it back when drawing:
score = |price - fit| / sigma band = fit +/- q * sigma
The band now scales with LOCAL DIFFICULTY — and note it is the MODEL'S difficulty, not the market's volatility. Related, but not the same thing, and the first one is what a prediction interval is actually about.
6. THE COVERAGE AUDIT. A guarantee you do not verify is just a claim.
TWO QUESTIONS ABOUT THE COVERAGE, AND THE PANEL ANSWERS BOTH
IS THE MISS REAL? That is a binomial z-test and it needs no tolerance at all. Each bar is a Bernoulli trial with p = nominal, so the standard error of the observed coverage is sqrt(p(1-p)/n).
IS THE MISS BIG ENOUGH TO CARE ABOUT? That is a judgement, and you set it.
These are NOT the same question, and conflating them is how a band gets waved through as "calibrated". Measured live on NIFTY futures: at n = 20,266 the standard error is 0.21 pp, so an empirical coverage of 89.1% against a nominal 90% is a 0.9 pp miss — FOUR STANDARD ERRORS. Unmistakably real. Arguably too small to trade differently. A 5 pp tolerance called that "calibrated", which was the headline row of the script asserting the one thing the script exists to verify, and asserting it wrongly.
The panel now reports the size of the miss, its significance, and a verdict that distinguishes "within sampling noise" from "real, but small" from "MISCALIBRATED — do not trust the band". You get to decide which of those matters to you, and you get the numbers to decide with.
AND THEN THE ROW NOBODY HAS: CONDITIONAL COVERAGE
Coverage 90.0% vs 90% n = 20,178
calm / normal / turbulent 96.4% · 90.1% · 83.2%
A single marginal number can read a perfect 90% while the interval covers 96% of quiet bars and 83% of violent ones. Ninety per cent is then the AVERAGE OF TWO ERRORS — it looks right while being wrong in both directions, and it is wrong in the direction that costs you money exactly when it costs you money.
Exact conditional coverage is provably impossible without strong assumptions. But you can always MEASURE it, and almost nobody does. Bars are split into calm, normal and turbulent thirds by the percentile rank of ATR, and coverage is scored inside each. If the three numbers fan apart, the band is not breathing — and the normalised score is what closes the gap.
Turn the normalised score off and watch those three fan out. That is the demonstration.
FADE OR FOLLOW? THE TOOL DOES NOT PRETEND TO KNOW
Price leaving a 90% interval is statistically unusual. Whether to FADE it (an outlier, so bet on reversion) or FOLLOW it (the model has broken, so bet on the new regime) is an EMPIRICAL question, and the honest answer is often neither.
So both are logged, both are graded, and BOTH ARE TESTED AGAINST EACH OTHER.
That last part matters more than it sounds. Knowing that fading beats an unconditional control, and that following also beats an unconditional control, does not answer the question a trader is actually asking at a band break — which of the two should I do? They are mutually exclusive responses to the SAME event. So they are run head to head with a Welch t-test, and the answer is allowed to be:
FADE or FOLLOW? NEITHER — the break does not tell you which
If the difference does not clear the noise, then on this instrument the break carries no directional information, and saying so IS the finding. A tool that cannot report its own failure is an advertisement, not a measurement.
And the chart agrees with the panel. An unproven direction is still drawn — it is arithmetic, and you may want it — but it is drawn MUTED and labelled "(not proven)". It used to print "Follow the break" in full colour while the panel directly beneath it said "neither proven". The paint has to agree with the code.
THE ANTI-BIAS GUARDS
ENTRY IS THE CLOSE, for the event and for the control alike. A band break is a SIGNAL, not a fill. Entering at the band — a better price — while the control enters at the close hands every signal a free head start and manufactures an edge out of nothing.
THE CONTROL IS DIRECTION-MATCHED. In a downtrend there are more break-downs than break-ups, so FOLLOW skews short and FADE skews long. A direction-skewed event set measured against a 50/50 control inherits the index drift for free and calls it an edge. Longs are compared only with control longs, shorts only with control shorts, and the control is blended back using the events' OWN direction mix.
IDENTICAL GEOMETRY. Every event and every control trade uses the same stop and the same R multiple, so the comparison is apples to apples.
Both barriers on one bar: the STOP is assumed first — conservative, and the only assumption that cannot flatter the result. Unresolved trades at the horizon are marked to market, not booked as losses. Nothing is marked proven below t = 1.96.
NON-REPAINT
The kernel is causal, the bandwidth is chosen on past error only, the interval is built from past residuals, and coverage is scored by asking whether the actual close landed inside the interval that was published BEFORE it. Everything is computed on confirmed bars. Nothing is drawn and then moved.
WHY THESE PARTS ARE ONE TOOL
The regression gives the trend. Without the interval, a band is a guess. Without model selection, the bandwidth is a knob you turn until you like the picture. Without the coverage audit, a conformal interval is an unverified promise. And without the signal calibration, "price left the band" is folklore. Each piece is worthless alone — which is exactly why they ship together.
DATA AND SCOPE
Any symbol, any timeframe. ATR-normalised throughout. No volume required.
EXPORTS (Data Window — consume from other scripts via input.source())
EXP_Fit, EXP_Upper, EXP_Lower, EXP_Bandwidth, EXP_Coverage, EXP_Miscal, EXP_Signal, EXP_Entry, EXP_Stop, EXP_Target
CONCEPT CREDIT
Nadaraya-Watson kernel regression — E. A. Nadaraya and G. S. Watson (1964). Its use for technical pattern recognition in finance — Andrew W. Lo, Harry Mamaysky and Jiang Wang, "Foundations of Technical Analysis", Journal of Finance 55(4), 2000. Conformal prediction — Vladimir Vovk, Alexander Gammerman and Glenn Shafer; the split/inductive form used here follows Papadopoulos et al. and Lei et al. Triple-barrier forward labelling — Marcos Lopez de Prado. Welch's t-test — B. L. Welch. ATR — J. Welles Wilder.
The causal-kernel implementation, the parallel bandwidth selection, the live coverage audit, the binomial calibration test and the fade-versus-follow head-to-head are the author's own. Clean-room implementation; no third-party Pine code is reused. Not affiliated with, nor endorsed by, any of the above.
HONESTY AND LIMITATIONS
Conformal coverage is guaranteed under EXCHANGEABILITY. Financial returns are NOT exchangeable — volatility clusters, regimes shift — so the guarantee is approximate in practice. THAT IS PRECISELY WHY THE COVERAGE IS AUDITED LIVE INSTEAD OF ASSUMED. When empirical coverage drifts from nominal you are watching the assumption break, in real time, and you should believe what you see rather than the label.
A prediction interval says where price is LIKELY TO BE. It says nothing about DIRECTION, and it is not a forecast. Coverage being correct does not make band breaks tradeable — those are two different claims, and the tool tests them separately for exactly that reason.
Calibration figures are IN-SAMPLE, with no costs or slippage, and use overlapping windows. A proven in-sample edge is NOT a guarantee out-of-sample. Nothing here predicts price.
DISCLAIMER
Research and educational tool only. NOT financial advice, NOT a recommendation, and NO guarantee of results. Entry, stop and target output is arithmetic, not advice. Trading carries risk of loss. Test out-of-sample and make your own decisions. The author accepts no liability for any use. 指标
