Research Archive
The Research Archive contains discontinued trading robot branches, inactive versions, earlier development stages and technical research records that are no longer part of the current active observation set.
Archived material is preserved because unsuccessful or discontinued systems remain useful. They show what was tested, which weaknesses were discovered, how research decisions were made, and why development moved in another direction.
Archive status does not mean that the underlying research data has been deleted. It means only that the version or branch is no longer part of active monitoring, scheduled execution or current development.
EURGBP V4
EURGBP V4 was a corridor-based EURGBP research branch tested through several live paper versions.
The branch progressed from the original V4 Control structure through V4.1 weekly modifications and finished with the V4.2.29 Selective SELL Groups research hypothesis.
Active development of this branch has been discontinued. Its logic history, weekly reports and final research decisions remain available as archive material.
EURGBP V5
EURGBP V5 was an independent EURGBP corridor research branch focused on parallel-corridor conditions and later selective SELL groups.
The branch progressed through weekly V5 versions and finished with the V5.29 Selective SELL Groups research hypothesis.
Active development of this branch has been discontinued. Its logic history, weekly reports and final research decisions remain available as archive material.
EURJPY V1 Corridor
EURJPY V1 was an independent corridor-based EURJPY research branch developed through broad impulse, parallel-corridor and selective-group testing.
The branch progressed through V1.27 and V1.28 and finished with the V1.29 Selective Groups research hypothesis.
Active development of this branch has been discontinued. Its logic history, weekly reports and final research decisions remain available as archive material.
FX Scalping A/B V1 and V1.1
FX Scalping A/B V1 and V1.1 were the earlier implementations of the Scalping research methodology.
These versions searched for short-term entries on predefined instruments. V1 used a global maximum of one open paper position, while V1.1 introduced a controlled re-entry structure.
The versions produced a meaningful number of profitable trades, but the overall risk structure was weak. Losing trades were too large relative to the gains and repeatedly outweighed the profitable results.
The methodology itself was not abandoned. Instead, development moved toward dynamic market selection, where the system first scans multiple instruments and then evaluates direction and possible entry conditions.
The A/B versions therefore remain archived as earlier development stages of the same Scalping methodology.
FX Dynamic Scalping Selector V1.3.2 Paper Funnel
FX Dynamic Scalping Selector V1.3.2 was an intermediate paper-funnel stage within the modernised Scalping methodology.
The version was technically safe and operated in PAPER_SIM_ONLY mode, but the Week 32 audit showed that its filtering structure was too restrictive.
During the reviewed period, the upstream selector produced 549 YES signals. The reconstructed V1.3.2 candidate set contained 128 rows, and only 41 closed paper trades were produced.
This showed that the funnel was removing too much of the available signal flow for efficient scalping research.
V1.3.2 was therefore used as the basis for the broader V1.4A Paper Funnel and was then disabled.
The version remains archived because it provided an important research result: excessive filtering can reduce a potentially useful signal stream to a sample that is too narrow for meaningful observation.
AUDJPY Follow Driver Research
The Follow Driver methodology studied whether movement in related driver instruments could provide an early indication of a delayed reaction in a follower instrument.
The AUDJPY implementation used:
- AUDUSD as a driver;
- USDJPY as a driver;
- AUDJPY as the follower instrument.
The main research versions were AUDJPY Follow Driver V0.2 and V0.3.
V0.2 initially produced a strong small-sample paper result and was later prepared as a controlled DEMO_ONLY version. V0.3 remained a parallel paper research version with a more aggressive timing structure.
The branch was eventually removed from active observation because the overall signal frequency remained too low to provide enough useful research data within a reasonable period.
Demo Duplicate-Signal Finding
During demo testing, an important technical issue was discovered: the same demo signal could generate more than one demo order.
Before the branch was archived, duplicate-signal protection was added. The protection checks whether the same symbol, direction, signal time and entry signal time have already produced a demo order.
If the same signal appears again, the duplicate is blocked before another order can be sent.
After the correction, the demo script passed validation and returned a safe status with orders_sent = NO.
Final Follow Driver Status
- AUDJPY Follow Driver V0.2 PAPER — disabled
- AUDJPY Follow Driver V0.3 PAPER — disabled
- AUDJPY Follow Driver V0.2 DEMO_ONLY — disabled
The branch remains valuable as research history because it documents both the follower-driver methodology and a practical execution-safety problem discovered during demo preparation.
AUDCAD MACD V Base
AUDCAD MACD V Base was tested as an unchanged MACD control line.
The robot remained active during the initial MACD observation period in order to collect independent statistics for AUDCAD without additional version filters.
Its scheduled task was later disabled and its working files were removed from the active observation area.
The recorded data remains preserved for future comparison or possible reactivation if a new research hypothesis provides a specific reason to revisit the line.
GBPUSD MACD Research
GBPUSD MACD was tested as part of the broader MACD observation programme.
One of the most useful findings from this line was technical rather than performance-related. A paper position could be opened and closed correctly while the closed-trade file failed to produce a valid R result.
This demonstrated that a robot cannot be evaluated reliably unless trade logging and state recording are complete.
GBPUSD MACD V30 was later removed from active monitoring, while the Base research history remains available for technical reference.
Other Inactive MACD Versions
This group contains MACD versions that were tested but are no longer part of the current active research priority set.
A version may be moved out of active observation when:
- its scheduled task has been disabled;
- the version does not provide a useful improvement over its control line;
- trade frequency is too low to justify continued monitoring;
- a technical issue prevents reliable evaluation;
- the active research set is deliberately reduced to clearer candidates.
Archived MACD versions remain part of the research history and may be reviewed again if a new hypothesis requires comparison with earlier behaviour.
Current Research Outside the Archive
The archive does not contain systems that are currently under active Week 33 observation.
The current external EA candidates are:
- ASQ Safe Scalping v1.20 — controlled DEMO_ONLY observation;
- Universal Breakout Study (UBS) — PAPER_SIM_ONLY observation.
These systems are documented separately because they are third-party EA candidates rather than internally developed research branches.
Open External EA Candidate Evaluation
Archive Principle
A robot does not need to be profitable to provide useful research information.
An archived system may document:
- a trading hypothesis that did not produce enough activity;
- a risk structure that proved unsuitable;
- a filtering system that was too restrictive;
- a logging or recovery weakness;
- a demo-execution safety problem;
- a research direction that was later replaced by a more practical approach.
Preserving these results helps prevent the same unsuccessful assumptions or technical mistakes from being repeated in later robot versions.
Archive Status
Archived research: retained for documentation and comparison
Scheduled execution: disabled unless a branch is explicitly reactivated
Live trading: not permitted
Risk Notice
Archived robot results, paper observations and demo observations do not guarantee future performance.
The Research Archive documents experimental trading-system development and should not be interpreted as financial advice, investment advice or a recommendation to trade live funds.