Manual Scalping versus EA

Can someone with some experience here tell me what the advantages are of a manual scalping system compared to an EA with regards to broker restrictions.

Assuming the EA does exactly what the manual system does albeit with perhaps not so good quotes.

Is it a case of what you are doing, rather than how you do it? eg. frequency of trades, duration of trades, pips per trade etc.

Thanks

I have no experience with this, but logic tells me that from the brokers side of things it is the actual trades they may find a problem with, not how they are being made.

But best to wait for someone who knows their stuff to chime in........

That's what I would have thought but I just wanted to check there wasn't any big disadvantage to converting a system to an EA.

Somewhere I thought I did see a comment that certain brokers ban scalping EAs. It may turn out that the comment was out of context and what they're actually doing is banning the trading style.

Hi guys. From my experience with manual scalp trading systems versus expert advisors. The advantage of trading the manual scalp trade method is if it is not a true scalp trading system you can cherry pick the best entries and exits based on your own analysis and discretion as opposed to pure mechanical automated scalp trade systems. Also for a scalp trade method to become an effective automated scalp trader you need to carefully formulate all rules of the system and test it to the outer limits before you can know how it will peform under all market conditions, so this includes testing it for longer than you may think is necessary. Of course there is less precision and execution time is longer when manual trading as opposed to using an automated bot. My personal preference is automated bot (EA) if the system is good and uses a great trade method. With this said scalp trading is one of the hardest trade style to pull off with consistent profits automated or not. As for brokers, well most MT4 brokers tend to dislike this kind of trading whether you use an EA or not. Just be careful to monitor trades closely if you do opt for the automated system because from my own experience some brokers will not fill you as price reaches your set take profit levels. When confronting my broker they blamed this on my scalp trading EA. These days spread is widened due to volatility and liquidity issues, we are told. It could be a good thing to base a trade method on trend following than it being blind, and also use a higher time frame for this than the time frame you execute orders on. Hope some of this makes sense and helps.

For me the most important thing to address is firstly, is the trade method viable; is it consistently profitable under the market conditions I wish to trade it with.

This can make a vast difference. Depending on your exit method of a trade, whether it is a standard take profit or some indicator or price action used to exit a trade. It all depends on risk tolerance as well. The more trades the higher the risk when scalp trading because most scalp methods is not true scalp trading. I would not want an ideal trade to last longer than 30mins with most closing out within 3 mins of opening but some scalp trade methods expose you for many hours, not true scalp trading if you ask me. All depends on your preference though because you may want to trail profit forward with the trend and only make one trade a day then. For this reason trading at a start of or during a trend is important, for speed and momentum of (price) each trade. For me then the answer to this question is to look for a good (trend) entry filter and to use a good exit rule in a scalp trade method, and to remember money management is paramount.

Sorry for the long post, hope it helps.

fxscalp, thanks for that. That all makes sense.

The attraction of an EA for me, assuming I can import all the features of my system, is it takes emotion out of the equation. Apart from turning it off during certain news event, I'd be happy to leave it to churn away.

I'm trying to filter the number of trades down to a one or two quality ones each day. Each trade could take anywhere from a minute to a few hours. Typically around 30 minutes.

I must admit I'm having problems with the actual definition of scalping as some brokers seem to differ quite a bit. I just spoke to one, who apart from totally evading every request for a specific definition on how they define scalp trades, even when I said I may only do one or two trade a day said it could be still be considered scalping. So I guess it's the duration of the trade. They wouldn't tell me the mininum though.

I agree about the difficulty. Scalping requires precise entries and exits and almost any EA will include at least 1 or 2 indicators and they lag the market badly. Entries can be easier but automated scalp exits will test your programming prowess!!

I have worked on many systems and have found it pretty difficult to get a EA to work on the 1 to 5 minute time frames. Along with the fact the scalp profits are minor and days worth of pips can be blown with a bad exit.

This is not to say it cant be done but automated 1 minutes scalps are VERY difficult. I've have have 80+% EA's that were still unprofitable when trying to scalp. This type of precision can ONLY be achieved with programs that compute their own price action signals and with a ECN style broker that gives your EA immediate feedback with flawless execution.

This is not a small task.

This isn't help at all by the almost useless MT Tester when it comes to scalping. It's very frustrating. Apart from using it for simple optimisation and debugging, forward testing seems the only way.

I had thought about getting some quality tick data for a year and developing a system outside MT4, being a programmer there's no problem there.

Then I think well why use an EA at all, just develop my own stand alone system but of course it gets messy with a trade interface and data feed. I do actually like MQL seeing it's based on my all time favourite language C.

true. 1 minute data is not good enough to backtest a true scalper. You'll have to forward test FOREVER to get profitable results. Not being able to test tick by tick is what stopped me from doing it. I dont develop anything that I cant backtest extensively!!!

I wish you luck. You can develop it and might get lucky and find the right settings but I know every EA I've ever made is developed in 2 stages - writing the EA and backtesting on Alpari databank 90% data until I've found the very best settings.