Making an EA

This journal will follow my attempts to create an E.A. that actually works.

Obviously I am daft enough to think that it is possible to trade automatically without me having to sit at the screen all day and night.

The journal will details my problems and my efforts to overcome them.

I have a basic E.A. that I have called Gemini.

In an ideal world I would like:

1. To be in the market most of the time, so that at least I get some entertainment value from this exercise.
2. I want to trade in micro lots of 0.01. So my risk is tiny.
3 I want a mixture of scalps and runners.
4.I do not want massive stop losses that wipe out the account when they are hit. In other words I want to be continually taking my profits and also my losses.
5. I want to trade GBPUSD, GBPJPY,USDJPY,CHFJPY,EURUSD,USD.
6. I want to be profitable without having to bite my nails- while I look at huge draw downs.

I have just launched Gemini on a new demo with Alpari.

I will now try to publish the initial results on a Trade Explorer at Forex Factory.


P.S. I am not interested in buying or selling anything here.

O.K my E.A. is up and running. It has already made a few trades.

I have also managed to publish the results on a trade explorer.

I am surprised just how much information the trade explorer actually publishes.

The first thing that I noticed was the huge stop losses and take profits. These are not the real stop losses. The E.A. shows the broker one set of stop losses and take profits and then works on a completely different set .

This Trade Explorer thing is great.

If anyone apart from me ever looks at this journal they will know by now that the huge stop losses are fake.

However, the small take profits are real.

When the E.A. sanctions a new trade it does two things:

1. It opens two identical trades.
2. One has a small take profit. The second trade is left open when the first one closes - but it's stop loss moves to break even.
3. So I have one profitable closed trade and one "runner" with a stop loss at break even. In snooker they call it "a shot to nothing". No doubt in forex I will get a bloody nose.

Well, the purpose of this journal is to detail the problems of a would be auto trader.

I only started yesterday and already I have had problems with my auto trading.

1. Last night the remote desk top that I have my Alpari MetaTrader on, let me down. They added a windows up date and then restarted the computer. This of course meant that when the computer restarted , my MetaTrader was not running. So my trades were not being managed.

2. Before going to work this morning I re-loaded MetaTrader. When I came home about half an hour ago I discovered that there had been a problem with the remote desk top. In fact it was not working and it seems to have totally shut it self off. Again my trades had not been managed.

My initial thoughts are that even if I had a great E.A. ( which is yet to be proved) it will never make a profit if the site it runs on is unreliable.

Oh well, things can only get better.

I had a look at my remote desk top earlier today and the EA was showing a small profit.

I tried to check it again about an hour ago but the remote desk top site would no longer allow me to sign in. I have sent them a help request.

I then checked my Trade Explorer on Forex Factory. The EA is now showing a draw down.

The only reason that I pay for a remote desk top service is that I want guaranteed internet conectivity and a reliable server.

Presumably other people who have considered trading solely with an EA have encountered similar reliability problems. My experience to date has been far from encouraging.

I have been denied access to my remote desk top for the last 24 hours.

Of course I am getting no replies to my requests for help.

I have now decided to abandon any attempt to use an EA on a remote desk - you have no control when things go wrong.

I have not enjoyed the experience, but at least I have learnt an important lesson.

I have now set up my EA on my computer at home.

Now I will be able to observe the EA in action.

I will be able to see if it is opening and closing trades according to my plan. By observing the EA in action I will also be able to spot improvements that can be made.

The results will hopefully appear on a Trade Explorer at the top of this thread.

As previously stated - I have given up the idea of trading with an EA on a remote desktop.

However I was alarmed to get an email from my provider. They said that my account had been hacked, and four user names unknown to me had been given access to the account.

No wonder I have been having so many problems trying to trade on a VPS !

If any of you are thinking of trading on a remote desk top - be careful.

I have removed my original EA as it wasn't doing very well.

I have replaced it with a simple trend based EA - which I will tweak from time to time.

I have also reduced the number of pairs traded to four.

I have closed down the EA until after the FOMC announcement.

No point in risking todays profits.

I will turn it back on when the markets settle down a bit.

I have made a ( conservative ) change to my EA.

In most cases it will only buy below the Daily Pivot or sell above the Daily Piovot.

It's no fun buying all the tops and selling all the bottoms.

This is intended to be the diary of someone (me ) trying to create a profitable EA that trades day and night with little or no input from me.

So far, I have not had much time to concentrate on making and improving an EA.

To date my main concerns have been with getting the EA to run smoothly - never mind profitably.

This morning I awoke to see a number of trades that were taking a pounding. It occurred to me that the EA should have closed them before things got so bad.

I removed the EA and re-loaded it. The re-loaded EA promptly closed the trades for a nasty loss.

If I had bought a superb EA for thousands of dollars -whats the point if normal people cannot expect the thing to run for months without technical problems.

Anyway - I have now decided to re-load my EA on a daily basis in case the previous loading has become tired or overloaded with data.

I changed the EA on Friday to make it much more conservative. I am now looking for a good trend, and I am keeping an eye on the Daily Pivot. It didn't trade on Friday or yesterday - so I was concerned that it wasn't working.

It seems to be entering and exiting the market o.k. today.

To begin with, my EA wasn't working properly so the profit /loss graph was not very pleasant viewing.

It now seems to be opening and closing trades as I would wish.

I am now off on holiday for a week - so I closed the EA and took my profits.

The graph looks quite good. I wonder how long that will last?

Still a lot of work to do.

In April this year I cobbled together an "indicator" of my own. I have been playing with it in my Live Account with Oanda. So far it seems to be doing o.k. All the trades in the live account are made manually.

I think that you can see the results of my live account if you click on my name

Now I am trying to incorporate this indicator into my EA.

I have virtually no mt4 programming skills which doesn't help much.

However I am finding that the more the EA emulates my live account trading style the better it performs.

Although I have not been totally happy with the performance of my EA - it has been making a profit - so I will try and take that as a postive.

The purpose of this journal is to keep me interested in the desire to trade profitably using an EA. Is this possible? - I really don't know.

However considering my very limited programming skills, the fact that I am in profit would seem to indicate that it is possible to trade with an EA and win.

544 trades so far and the EA is still in profit.

One big draw down so far - this was due to an error in the coding.

Once the error had been fixed everything has been fine.

I have turned off my EA today.

I do not want to risk any open trades ahead of the Greek Elections.

Unlike some people my indicator does believe in Technical Analysis.

It had a very good day today ( so far ) buying the GBPUSD.

This demo account for my E.A. was opened on 30th April this year.

So far my EA has opened and closed 502 trades.
The win rate is 61.4%

I started with a 5000$ account but due to the tiny lot size I tend to use virtually none of the available margin. In real life I think that I could operate this EA on a live account of 1000$ or less.

Each trade has a lot size of 0.01

Given that the lot size is miniscule I think that the profit of 381$ is respectable.

Since I found a stupid error in the coding a few weeks ago the E.A.s Draw Downs have been well managed by the E.A.

We have been running for less than two months - so it's too early to say that this experiment has worked. But so far so good.

Today my EA made 160 pips ( so far ).

So far this month it has made 1770 pips.

I'm pleased with that.

I see that according to my explorer it was another good day today.

Profit for the day was 90 pips.
Profit for the week is 250 pips.

The EA must be doing something right.