Disliked{quote} I don't loose money in trading. I am profitable 7 years in a row & i 'don't have any "evil" ideeas...we talking about pentesting the MT4 security in this thread.If i go over and disassemble the terminal.exe in memory while running + sniffing all the data packets before being encrypted and sent to broker ... who know what else can find.But i have no interest in doing so.I only exposed possible security problems. I could easily program an evil robot to read and anticipate all your EA's future orders creating and sending a virtual order to...Ignored
- #32
- Edited 8:18am May 21, 2019 8:04am | Edited 8:18am
- Joined Feb 2012 | Status: Trader | 3,928 Posts
TSLA: 30-Dec-22 123.18 // 16-Jun-23 260.54 **111.51%**
- #37
- Aug 28, 2019 1:34am Aug 28, 2019 1:34am
- Joined Jul 2016 | Status: Trader | 2,534 Posts
Thanks for the good laugh fellas. Here, I found the solution to stop all the brokers from stealing the valuable codez from your mind/EA/airgapped PC.
- #38
- Edited 9:34pm Aug 30, 2019 1:55pm | Edited 9:34pm
- Joined Sep 2006 | Status: ... | 8,113 Posts
Thanks for the good laugh fellas.
Some possible reasons:
1. There are easier and risk-free ways for brokers to make money without any effort: spread, commissions, swap differential. And they can 'A-book' the novices and losers, and simply run their own automated processes to trade against them. Or they can cheat their clients by using VDPs, running their stops, widening spreads, slippage, etc.
2. The vast majority (99%+?) of retail EAs are ineffectual, locating little better than 50/50 bias, and will lose money if used for long enough. Many are profitable for a while because they either use MM that's ultimately unsustainable (grids, martingale variants, loss avoidance, recovery techniques, etc), or due to statistical fluctuation (read this), but the law of large numbers will eventually catch up with them. These types of EA are a dime a dozen and can be found almost anywhere. There are no hidden secrets: no matter how 'good' a system appears to be, trading involves losses when market conditions are unfavorable, which equates to unnecessary risk for a broker (and there's no absolute guarantee that what is profitable now will continue to retain its effectiveness over the coming months/years). And there's always the additional risk that the code is buggy (and its random effect has approximately 50% chance of improving the strategy!)
3. Some EAs are manually turned on/off (e.g. around high impact news), or otherwise involve some form of user intervention (e.g. deciding which instrument to trade**), hence the trader's overall IP can never be completely discovered. Even if this isn't the case for a given EA, the broker has no easy way of knowing that.
**Example: the owner currently runs his EA on pair(s) that proved most profitable during a backtest performed 6 months ago. Then he re-tests today, and the new data causes a different result, so henceforth he runs it on a different basket of pairs. The broker has no way of knowing when or how the owner performs this determination, or if such a background process even exists.
4. Some EAs are a work in progress, i.e. rules will be added or changed as the owner's knowledge increases, or his perception of 'changing' market conditions dictates. Going forward, these rule changes might not necessarily improve the EA's performance.
5. The 'technical' personnel employed by brokers are: (i) not knowledgeable traders (if they were capable of making a living as a trader, why are they working 9-to-5 for a brokerage?), and (ii) engineers who understand how to navigate Windows, and have been trained well enough in operating both the broker and client sides of MT4/5, but have little specialist programming knowledge.
6. I'm not sure how the current MT4 operates but, prior to build 600, only your orders were sent to the broker; your .ex4 remained on your computer. Hence the broker would have to hack your computer to steal it. Moreover, the .mq4 code wasn't stored in the EA; that's why the best that decompilation could achieve looked something like the assembly language used by computers back in the 1960s; a LOT of effort would be required, even for an experienced programmer, to reverse engineer the trading rules. Here's a snippet from a decompiled .ex4:
In summary then, brokers aren't interested in your IP, because even if a retail holy grail existed and you happened to own it, it wouldn't be worth their effort and cost to unravel it.
