【Exposure】Expose the Lies of EA Backtesting! Don't Be Fooled by Bogus Verification!
Here are BASARA's EAs on sale
https://www.gogojungle.co.jp/
[Expose] Exposing the Lies of EA Backtests! Don’t Be Fooled by Deceptive Verifications
“This EA has never collapsed in 20 years!” “Incredible win rate of 98% in backtests!”
…Have you ever been tempted to click after seeing such slogans?
But can you really trust those backtests?
To start with the conclusion: backtests can be easily “fabricated.” And many backtest results flooding the market arelikely crafted with nearly fraudulent specifications.
In this article,we will thoroughly explain the tricks used to fake them and how to spot them.
1. Backtest period is too short to be trustworthy
When evaluating an EA,the backtest periodis extremely important.
Nevertheless, there are many cases where results from less than a year, or even just a few months, are公開 as “excellent.”
That simply means they are showing only a period where the EA happened to win.
FX markets constantly change: ranging, trending, rapid rises and falls, economic releases, flash crashes… Whether a logic can withstand all of these must be proven bylong-term testing of at least 10 years.
2. Realistic spreads are not used
Spread settings also require scrutiny.
In MT4 backtests, setting the spread to “10” means 1.0 pips, but setting it to “2” results injust 0.2 pips.
Even for USD/JPY or EUR/USD, having a constant 0.2-pip spread is not realistic in the real marketand is impossible.
Spreads are especially impactful forsingle-position EAs.
Hedging EAs can be offset by multiple positions, but single-position EAs cansee large changes in profit with even a 1-pip difference.
3. Look at the modeling quality! Do not trust anything below 99.90%
The “Modeling Quality” shown by MT4 is also a critical indicator.
| Backtest Method | Modeling Quality | Reliability |
|---|---|---|
| 1-minute + all ticks | Below 90% | ✕✕✕ |
| 5-minute | Below 90% | ✕✕ |
| 1-minute (n/a) | Unclear | ✕✕✕✕✕ |
| TDS (Tick Data Suite) | 99.90% | ◎◎◎ |
If the modeling quality is below 99.90%, that backtest has little value.
In particular, backtests that use freely downloadable historical data raise concerns about data accuracy..
In fact, in cases the author tested,a grid/server EA that would be immediately liquidated in real conditions ran for 20 years without failing in free data.
This is more of an “informational record” than a demonstration of real capability.
4. The TDS “trick” can beautify results any way you want
“TDS (Tick Data Suite)” is famous for enabling backtests with realistic accuracy, but it can also beabused.
TDS allows the following customizations:
-
Spread: can be set to 0.1 pips
-
Swap: both buy and sell can be set to positive
-
Fees: can be set to zero
In other words, an EA that would normally lose can be made to produce a steadily rising equity curve by relaxing the conditionsand easy adjustments.
5. Existence of “don’t lose” cheat scripts
Some developers even embed scripts into MT4 Strategy Tester to intentionally exclude losing trades..
This is completely within the realm of fraud..
“Stops are hit but losses aren’t realized”, or “trades execute at prices that clearly shouldn’t,”unusual backtest resultshave hidden motives.
6. What backtests can be trusted?
So,which backtests are reliable?
Key points:
-
Use TDS (99.90% modeling quality)
-
Spreads and swaps reflect reality
-
Long-term testing over 10–20 years
-
Monthly and yearly profit/loss trends are visible
-
Forward (live) performance aligns closely
If these are missing, backtests should be regarded asfor reference only.
Summary: Backtests can be made to look better, so be skeptical
EA performance can be made to look “better” through backtestsand that is common.
Many sellers and affiliates use glossy numbers to market products.
However,if you pick an EA that loses in real trading, you will lose.
Therefore, always verifyspread, modeling quality, period, data source, and validation methodto avoid being fooled by numbers alone.
Final thoughts
Truly excellent EAs deliver results without excessive backtesting manipulationand that ability separates winners from losers.
I hope this article helps you pick an EA and assess backtests more effectively.
?Choose wisely, and do not be fooled!
Here are BASARA's EAs on sale