← Explore all 5 free MT5 indicators
Forex Top Team / Kernel Envelope and Mean Reversion / MT5
Reversion Kernel M5
Explore the return toward the mean.
How far has price moved from its local mean? Inspect a smoothed price curve, its volatility envelope and closed-candle re-entry candidates, with a virtual model that makes its assumptions visible.
Free for approved referral accounts · Account-specific EX5 · English user guide · Manual analysis

Watch the stretch as well as the direction
Reversion Kernel M5 helps you investigate potential returns toward a local price mean. A one-sided Gaussian kernel smooths price, while an envelope frames its recent variation. Confirmation filters screen candidates rather than assuming every stretched move will reverse.
A smoothed reference for local price.
Variation around the mean, with an ATR width floor.
A closed candle returning inside the envelope.
These describe the analysis process, not current market readings. M5 is the intended starting timeframe. On another timeframe, calculations use that chart’s timeframe and the panel displays an M5 recommendation.
How to read the chart
| Display | Meaning |
|---|---|
| Gold line | The one-sided Gaussian kernel mean. |
| Blue lines | Upper and lower envelope boundaries. |
| Green / red arrow | A confirmed buy / sell candidate. |
| White horizontal line | Entry of the active virtual trade. |
| Red horizontal line | Current stop of the active virtual trade. |
| Three green horizontal lines | The virtual model’s TP1, TP2 and TP3. |
| Performance panel | Rolling historical OHLC simulation statistics, not real account results. |
The envelope uses weighted price standard deviation around the mean, with an ATR-based minimum width. Horizontal entry, stop and target lines appear only while the virtual model has an active trade. They do not represent your broker positions or actual orders.
What qualifies a re-entry candidate?
Buy: a return from below the lower envelope
The previous closed candle must close below its lower envelope. The signal candle then closes back inside the envelope while remaining below the kernel mean. By default, the previous RSI is at or below 35, RSI rises on the signal candle and that candle is bullish.
Sell: a return from above the upper envelope
The previous closed candle must close above its upper envelope. The signal candle then closes back inside the envelope while remaining above the mean. By default, the previous RSI is at or above 65, RSI falls on the signal candle and that candle is bearish.
Trend and spacing filters
With default filters, ADX must be at or below 28 and the mean’s average absolute slope over five candles must not exceed 0.25 ATR per candle. Another fully confirmed candidate in the preceding three candles suppresses a nearby candidate; suppressed candidates also count in this spacing check.
Closed-candle timing: the latest closed candle is checked after a new candle starts; the unfinished candle never receives an arrow. The current envelope can move with price. The smoother uses the evaluated candle and older data, not future candles. History corrections, changed inputs and the rolling display window can alter displayed results.
A candidate is chart information, not an instruction to enter every arrow. Arrows and alerts may appear while the virtual model is already holding a trade and cannot accept another model entry.
Understand the virtual trade model
The model allows one active virtual trade at a time. If flat at the opening, it enters at the next candle’s opening price following a signal. Signals while a model position is already active are skipped; an opposite signal does not automatically reverse the position.
Initial risk distance is the signal candle’s ATR multiplied by 1.5 by default, with a minimum distance of two price ticks. That distance defines 1R for the model trade.
| Model event | Default behavior |
|---|---|
| Initial stop | One initial risk distance against the entry direction. |
| TP1 | 0.75R; closes one third of the original virtual position. |
| TP2 | 1.50R; closes another third. |
| TP3 | 2.25R; closes the final third. |
| Break-even after TP1 | Moves the remaining stop to entry from the following candle when enabled. |
| Holding limit | Remaining position exits at the close of the 60th holding candle. |
Reaching all three default targets produces +1.50R before costs for the whole original position, because the exits are split into thirds. An entry-price stop does not cover trading costs.
If OHLC data cannot determine whether a stop or target was touched first inside a candle, the model applies the stop first. Adverse opening gaps beyond the stop use the opening price, while favorable target gaps use the target price. Intrabar hits are processed after that candle closes.
The model does not trade your account. Modeled entries, partial exits and break-even changes are internal calculations. Reference levels are not normalized for actual broker order placement or lot constraints. No real orders or protective stops are created.
Read simulation statistics with their assumptions
The panel shows closed virtual trades, win/loss/flat counts, total and average R, profit factor, closed-trade drawdown, target hits, stop exits, timeouts and skipped signals.
- Historical simulation: statistics come from the chart’s OHLC data and the model’s assumptions.
- Default GROSS results: the round-trip cost input is zero by default.
- Closed trades only: active trades are excluded from closed-trade counts, Net, Win, PF, Avg and Closed DD.
- Closed DD: drawdown is measured at full trade closures and excludes floating drawdown.
- Rolling window: the model is rebuilt over the configured history window, starting flat. It is not a persistent live trade ledger.
Simulated statistics do not predict live performance. Variable spreads, swap, ordinary slippage, actual fills, minimum commissions per partial exit and lot rounding are not modeled. The optional fixed round-trip cost is entered in points, not pips, and is deducted once per fully closed trade.
Key default settings
| Input | Default | Purpose |
|---|---|---|
| InpKernelBars / InpBandwidth | 80 / 16.0 | Kernel length and smoothing bandwidth. |
| InpEnvelopeDeviation / InpMinimumWidthATR | 2.0 / 1.0 | Weighted deviation multiplier and minimum width on each side. |
| InpUseRSI / InpRSIPeriod | true / 14 | RSI confirmation. |
| InpRSIBuy / InpRSISell | 35 / 65 | Previous candle’s RSI boundaries. |
| InpUseADX / InpADXPeriod / InpMaxADX | true / 14 / 28 | Maximum ADX filter. |
| InpUseSlope / InpSlopeBars / InpMaxSlopeATR | true / 5 / 0.25 | Mean-slope limit. |
| InpCandleConfirmation / InpCooldownBars | true / 3 | Matching candle direction and candidate spacing. |
| InpATRPeriod / InpStopATR | 14 / 1.5 | ATR and initial model risk distance. |
| InpTP1R / InpTP2R / InpTP3R | 0.75 / 1.50 / 2.25 | Three-stage model targets. |
| InpBreakEvenAfterTP1 / InpMaxHoldingBars | true / 60 | Remaining stop adjustment and model holding limit. |
| InpRoundTripCostPoints / InpHistoryBars | 0.0 / 3000 | Fixed modeled cost and rolling history window. |
| Popup / push / sound alerts | Off | Optional notifications for confirmed candidates. |
Target inputs must satisfy 0 < TP1R < TP2R < TP3R. Changing filters, bandwidth or costs can materially change the modeled results. Looser filters do not establish better performance.
Get your free account-specific copy
Choose any broker listed on our broker guide that offers an eligible MT5 account for your location through the listed referral route. HFM is our first choice, but it is not required.
Our preferred route: HFM
HFM is Forex Top Team’s primary recommendation for eligible users. Other brokers listed on our guide may also qualify. Confirm your residency eligibility, MT5 availability and the chosen broker’s referral requirements before registering.
- Open an eligible MT5 account. Use your chosen broker’s referral link or code from our guide. Existing clients should confirm additional-account referral attribution with the broker.
- Send your activation request. Include Reversion Kernel M5, your MT5 trading account number, broker, exact MT5 server name and delivery email. Do not send passwords.
- Receive your personal EX5. After eligibility is verified, we compile and send your account-specific file with an English user guide.
Our top pick: open an HFM MT5 account ↗
Suggested subject: Reversion Kernel M5 — Account Activation
Use your MT5 trading login, not your broker profile ID or MQL5 community ID. Activation is issued manually; the local account lock does not automatically verify referral attribution. The approved login is embedded in the compiled file and cannot be changed through Inputs. A different account requires a separately issued copy. Some builds may also check the broker server.
Installation and first use
- Log in to your approved MT5 account.
- Select File → Open Data Folder.
- Copy Reversion_Kernel_M5.ex5 into MQL5 → Indicators.
- Refresh Indicators in Navigator, open an M5 chart and attach Reversion_Kernel_M5.
- Read the English guide and review the confirmation, target, cost and history settings.
Allow sufficient history for kernel and indicator warm-up. For push notifications, configure your MetaQuotes ID under Tools → Options → Notifications, then enable the indicator’s push input. Initial attachment does not announce historical signals. Alerts refer to confirmed candidates, including candidates skipped by the single-position model.
This independent MT5 indicator supports manual analysis. It does not open, modify or close orders and does not require DLL or automated trading permission for its calculations.
Frequently asked questions
Is HFM required?
No. An eligible MT5 account opened through any referral route listed on our broker guide may qualify. HFM is our preferred option.
Does the virtual model manage my real trades?
No. All modeled entries, partial exits and stop changes are internal simulations. You remain responsible for real trade management.
Does a target at 2.25R mean the whole model earns 2.25R?
No. The original position exits in thirds. Reaching all three default targets totals +1.50R before costs.
Are the panel results from a real account?
No. They are rolling historical OHLC simulations. With default zero costs, results are gross. Floating drawdown and several execution costs are not included.
Why can an arrow appear without a new model trade?
The raw signal and the virtual model are separate. If the model is already holding a trade, it skips the new candidate rather than adding or reversing a position.
Why are no horizontal trade levels visible?
The virtual model may be flat, or InpShowTradeLevels may be disabled. Levels describe an active model position, not your broker trades.
Why do model exits update after the candle closes?
The simulation processes closed OHLC candles. Developing intrabar target and stop touches are not processed until the candle is closed.
Can I use my file on another account?
No. Request another account-specific file after that account’s eligibility is confirmed. If your copy includes a server restriction, that server must also match.
Does it guarantee profitable trades?
No. Signals can fail. No verified profitability or win-rate claim is made. Compilation checks do not establish live chart behavior, successful account authorization or trading performance.
Trading risk: Forex and CFD trading can produce substantial losses. This indicator is a decision-support tool, not a profit guarantee. Simulated statistics do not predict future returns. You remain responsible for entries, position sizing, stops, exits and account monitoring. Review the broker’s terms and your own risk limits before trading.