Treasury Auction Inventory Fade
- Experts
-
Donizeti Aparecido Alves Ferreira Junior
Software developer and founder of D7 Bots, building algorithmic trading tools for MetaTrader 5. - Version: 1.3
- Activations: 5
TreasuryAuctionInventoryFade is an MT5 Expert Advisor for GBPUSD with an internal M30 decision timeframe. It is an event-driven strategy that measures GBPUSD displacement before selected U.S. Treasury 3-year and 10-year note auctions and trades a fade at the 13:00 New York event bar.
The trading symbol is configurable through InpTradeSymbol , with GBPUSD as the default. Strategy data, ATR, quotes, symbol specifications, position checks, sizing and orders use that configured target on M30. The chart that hosts the EA is not a substitute market or signal source.
Audited default logic
- Target market: GBPUSD, internal M30.
- Economic event: U.S. Treasury 3-year and 10-year note auction dates.
- Event clock: 13:00 America/New_York.
- Pressure window: 120 minutes in frozen G04.
- Direction: fade the signed pre-event displacement; positive GBPUSD displacement produces SELL, negative displacement produces BUY.
- ATR: 20 completed M30 bars.
- Stop: structural pressure-window extreme plus 0.10 ATR, with a minimum stop distance of 0.50 ATR.
- R target: disabled.
- Time exit: 120 minutes after entry.
- Session fail-safe: 15:55 New York.
- Spread guard: maximum 0.10 ATR.
- Trailing and breakeven: disabled.
- Regime overlay: NONE.
- Magic number: 26090672.
- Default risk: 1.00% of equity. The interface permits values through 10.00%, including 5–10%, but all historical evidence below used 1.00%. Higher risk can materially increase loss and drawdown.
Historical Strategy Tester evidence
The commercial figures below are historical MetaTrader 5 Strategy Tester simulations from the frozen scientific source and G04 SET. They used MetaQuotes-Demo, model 1, virtual initial deposit USD 10,000, and a native GBPUSD M30 host.
In-sample: 2020-01-01 through 2024-12-31
- Net profit: USD +1,817.05
- Trades: 143
- Profit factor: 1.26
- Maximum balance drawdown: 10.14%
- Observed frequency: 2.349 trades per 30 days
Annual net was not uniformly positive:
- 2020: USD +114.50
- 2021: USD -17.91
- 2022: USD +1,044.77
- 2023: USD +1,701.72
- 2024: USD -1,026.03
Out-of-sample test interval: 2025-01-01 through 2026-10-03
- Net profit: USD +256.80
- Trades: 21
- Profit factor: 1.26
- Maximum balance drawdown: 4.28%
- Observed frequency over the nominal interval: 0.984 trades per 30 days
The retained OOS ledger shows:
- 2025: USD +256.80, 21 closed trades.
- 2026 through October 3: USD 0.00, 0 trades.
- Last retained OOS close: 2025-12-09 22:00 server time.
The zero-trade 2026 segment is not hidden. The scientific calendar ended on 2025-12-09, so the approved historical OOS report contains no 2026 auction trades.
Current calendar coverage
For product maintenance, release 1.03 preserves every historical embedded date and adds only future dates after the approved OOS endpoint of 2026-10-03. As of 2026-10-04, the official U.S. Treasury Tentative Auction Schedule lists the following same-type auctions:
- 2026-10-06 — 3-Year Note
- 2026-10-07 — 10-Year Note reopening
- 2026-11-09 — 3-Year Note
- 2026-11-10 — 10-Year Note
- 2026-12-07 — 3-Year Note
- 2026-12-08 — 10-Year Note reopening
- 2027-01-11 — 3-Year Note
- 2027-01-12 — 10-Year Note reopening
These eight rows are official tentative schedule dates, not P&L-selected dates and not new backtest evidence. This maintenance did not retrieve a security-level offering announcement confirming the October 3-year/10-year rows, so it does not label any of the eight rows as individually announced. Treasury can change tentative dates; a later security-specific offering announcement supersedes a conflicting tentative date.
The embedded calendar is static rather than downloaded at runtime. Coverage after 2027-01-12 is not inferred. The calendar must be maintained from primary Treasury sources as newer schedules and offering announcements become available.
Host-chart timing boundary
The historical performance numbers above belong to the native GBPUSD M30 host. A different host chart can generate OnTick callbacks at slightly different instants. The strategy still reads and trades GBPUSD M30, but exact execution price/time can differ because the callback that notices a new target bar may arrive on a different host tick. Cross-host conformance therefore checks target symbol, event date, direction, trade count and execution path; it is not assumed to reproduce identical P&L tick for tick.
There is no fallback asset, artificial trade, tester-only economic branch or alternate validation signal.
Costs and execution limits
The retained deal ledgers record USD 0.00 commission and USD 0.00 swap in the two commercial runs. Parsed metrics do not expose a separate fee total, and spread is not separately decomposed from executed prices. No claim is made that another broker will reproduce those costs.
The evidence is not Vantage evidence, not real-tick validation and not live profit. Contract specification, spread, tick value, stops/freeze levels, volume steps, margin, server clock, trading sessions and symbol naming can change execution.
Historical simulation does not guarantee future performance.
