Weekend Gap Fade for MT5

Weekend Gap Fade: H1 cost-research campaign

Profitability is not established. This is 20 frozen Strategy Tester cases across ten independently reset USD 10,000 EURUSD segments, not one compounded curve or a holdout.

Rules and configuration

The public original TFPriceResearch_Sessions.ex5 remains the corrected-session-0 download. Campaign figures use the separately compiled tester-only research wrapper, which adds accounting exports but does not change strategy signals. It is limited to EURUSD.TFQ1 H1, mode 3, variant 0, magic 26092101 and session 0.

On a new H1 bar, the completed Monday bar must have broker hour 00–03 and follow the weekend; entry normally occurs from broker hour 01–04. The strategy takes the most recent Friday H1 close within four days, fades a gap of at least 0.50 completed ATR(14), sells a gap up and buys a gap down. Target is that Friday close and is mandatory; the protective stop is at least 1.5 ATR, widened for broker minimum distance. Only one owned position may be open. The RAM daily-equity brake is 1.5%; its baseline resets on restart, so restart behavior is unverified.

It rejects spread above 2.0 pips, sizes at 0.25% equity risk, and includes the source USD 7 round-trip-per-lot estimate only in stop-risk sizing. Bid/ask and execution can differ from the signal quote. Positions close through SL or TP, or at the first tick after 24 hours. Tester-end closes are separate from ordinary exits. Base models USD 3.50 per lot-side (USD 7 round trip) with no added delay; stress models USD 7 per lot-side (USD 14 round trip) with 200 ms delay. These withdrawals are separate from the USD 7 sizing estimate and native swap.

The frozen windows span 2018–2026, using genuine released bid/ask ticks with documented gaps. Broker time uses UTC+2 in winter and UTC+3 during US daylight-saving time. The 2020 segment starts 7 January, excluding the first Monday because its preceding Friday falls in an approved data exclusion.

Each scenario closed 75 positions: 50 at take-profit, 23 at stop-loss and two through the 24-hour exit. There were no tester-end liquidations. The four time-exit observations across both scenarios occurred exactly 86,400 logged seconds after entry; subsecond timing is not established.

Frozen observations

Aggregate interpretation: base: price P/L USD -6.79, fees USD 52.15, PF 0.8930, swap USD -0.29; exits TP/SL/other 50/23/2; stress: price P/L USD -22.98, fees USD 104.30, PF 0.7780, swap USD -0.29; exits TP/SL/other 50/23/2. Price P/L is already negative in both scenarios and modeled fees further worsen the result. These independently reset observations do not identify a causal reason or qualify profitability.

The public EX5 and wrapper matched in one selected February 2024 zero-fee trade. That limited comparison does not prove full multiyear equivalence. Historical financing, complete ATR/Friday-target reconstruction, risk-volume calculations, restart handling and live execution remain unverified. The RAM daily brake is not persistent. No profitability badge is awarded.

Download the public original session-0 ZIP · Frozen cost facts (JSON). No account orders are part of these tester-only instructions.

Earlier single-trade functional check

Updated 23 September 2026. Functionality demonstrated; profitability after costs remains unverified.

The older selected preset limited entries to London 08:00–11:00, while this strategy normally evaluates its first Monday H1 gap near the broker week opening. The main download and preset editor now use CampaignSession=0. The EA binary is unchanged; load selected-test.set to select mode 3, variant 0.

Entry, stop and exit rules

On a new H1 bar, the most recent completed bar must be Monday broker hour 00–03 and the preceding bar must not be Monday. Compare that Monday bar’s open with the latest Friday H1 close. Fade an opening gap of at least 0.50 × completed ATR(14): sell a gap up or buy a gap down. The mandatory target is the Friday close, which must remain on the profitable side of the current entry beyond the broker minimum distance.

Stop distance is at least 1.5 × ATR(14), widened for broker minimums. Size from 0.25% equity risk, with the configured 7-per-lot account-currency fee estimate included in sizing; reject below-minimum volume. Spread ceiling: 2 pips. A 1.5% daily equity brake blocks new entries and resets on restart. Exit by SL, TP or the first tick at least 24 hours after entry. This tester-only EA has no matching contract among the 31 core library specifications: this page documents its own distinct gap-fade strategy.

Paired native test on approved real ticks

EURUSD.TFQ1 H1, February 2024, USD 10,000, 1:100, no added commission or delay: both cases processed 1,706,310 real ticks and 504 H1 bars. Original session 1: one signal, zero orders, one aggregate block. Corrected session 0: one 0.15-lot SELL at broker 26 February 01:00, followed by an ordinary TP exit at 02:22:43; no forced exit. The only changed EA input was the session filter.

Price profit was USD 8.40, commission and swap were zero. The fee estimate used in lot sizing is not a charged commission. One positive trade is not an after-cost profitability qualification. ATR recurrence, volume sizing, losing trades and the natural 24-hour exit were not independently validated by these two cases. Session causality follows the source rules and paired outcomes; the original log does not record individual veto branches.

Download EA with corrected research preset · Versioned functional test status · Original session-1 archive for historical reports

The historical graphs and old report tables on this page belong to the original session-1 configuration and incomplete broker history. They do not show the corrected preset’s performance. The corrected package retains the old preset separately for comparison.