The purpose of the Algo Profit Booster
tool is to try to boost the profit-performance of any Expert Advisor
that has at least one optimizable input variable. It applies equally
well to Expert Advisors are highly profitable on out-of-sample data,
as well as ones that perform poorly. Both developers and users of
Expert Advisors alike might benefit from this tool.
Typical algorithmic trading involves a set of input parameter
values, often pre-optimized, that remain static during trading in
real-time even when underlying market characteristics may be
changing. This tool introduces a paradigm shift. It works by
dynamically adapting and changing the parameter values, while trading
in real time, using a intelligence knowledge base モlearnedヤ
earlier. No more trading with fixed input parameter values.
This is a multi-part software package of which only the MQL
library element is included in this product. The rest of the package
is free and must be downloaded. A link to the download page will be provided after purchase. A complete User Guide is included in
the download package.
User doesn't need to write any MQL code to use the APIs provided
by this library. All the necessary code is generated by an MQL code
generator utility included in the download package.
This is a versatile tool that can be applied to any trading algo
(a.k.a Expert Advisor or Robot) with a strong potential for boosting
trading performance in one or more of the following aspects:
To achieve its objective, the expert advisor code is augmented
with additional code to analyze and capture essential intelligence
during analysis and learning phases.
This tool is available in 2 different feature set,
A Single-Instance (SI) version, and
A Multi-Instance (MI) version
Both feature set options are identical in functionality, only
differs on how many instances (charts) can be active at a time when
trading with it. The Single-Instance version can be active with one
Expert Advisor on one chart at a time, whereas the Multi-Instance
version has no restrictions.
This particular product is for MetaTrader 5 and for the Single-Instance version. After purchase please send me a chat message or an email to the "Contact Us" link at expertprofitbooster.com and I will provide you with a link to the download page for the companion package.
Note: If APB_EA_Booster_SI.ex5
file is found in MQL5/Experts/Market folder, it must be manually
moved to the MQL5/Libraries folder.
The following result is from an actual
test after applying the booster tool on モSimple Moving Average
crossoverヤ Expert Advisor on EURUSD, Daily.
As expected, the base Expert (in black
color) fails badly on out-of-sample data but the booster produces profitable
results. Following is the test setup,
Base Expert:
Simple Moving Average crossover (Fast moving average and Slow moving
average)
Symbol:
EURUSD, Daily bar, Spread=10 pips, Fixed Lot Size=1
Money
management: None
Learning
Algo_prefix: -1, Trading option: 4
In-sample
dates: start-date=2000.01.01, end-date=2015.08.31
Out-of-sample
dates: start-date=2015.09.01, end-date=2018.06.30
Input
parameter optimization setup: start=3, stop=21, step=1, for both
fast and slow average lengths
Note that even though the booster does
well in the in-sample period also, this is of no significance. What
matters is the out-of-sample performance. Refer to the User Guide for
more details.
Even though the example above is from
an actual test, it is presented here for illustration purpose only.
It is not a recommendation for any specific Expert Advisor.