Rekon-X Auto Bots Expert Advisors for Forex Trading

So, let me get this straight. Its only available on demo for limited time and you want FF users to test and report bugs and shit so you can later sell/rent this?

Ha ha, not like that buddy

Hi

Why not fro MT4 as many of the members still use MT4 and not MT5?

Count me in +1...

For mt4 the coding is different and the libraries i have used are not available. May be later we can check for MT4.

Then how is it? You're quite unclear about it... If its available then share it here and hopefully working on demo and live

Yes, please have patience. BTW, please refrain from doing random comments.

Alright, hope all set!

In case you have not read my past work yet, please read them from my profile page. It will be helpful for you to understand whom you are following and it will easy for me to demonstrate the work done!

Majority, still wonder why would someone share his/her work openly - so my answer is, there are many who shares there work openly and later regret why they have shared such good work for nothing! I have follwed them, studied them from last several years so I am inspired by their work and method of open sourcing at the same time I know that they are not happy with the response they have received for their hard work. That is the reason, I prefer being private while sharing my work openly.


Enough preface for now, let's get back to the work details.


A. Time Series Analysis Method and why we need that?

Time series analysis, means past history study? There is no merit doing that. A new days starts with new perspective nothing to do with what happened earlier.

Is this so? Then, what about them who trades on 5, 35/50, 100, 200 SMMA/Moving averages? And does quite well. Simply following MA strategy plotted on Monthly to 1 minutes visual can yield you enough profit so that you need not to wait for 3 years to tell it is working!

Let's go one step back and break it down more.

I need USD 100 profit out of 300 for September'17 month, can I make a strategy in August'17 to make that happen. Or I need my capital of USD 300 grow every month by 5% so a 15 USD profit per month but my 300 dollar should not go below 270 dollar at any moment.


The above two are completely two different scenario, which one you like will tell you who you are and what you want?

There can be multiple scenarios too, but let's not mix them here.

We are going to focus on the method A which is, I will have a fixed dollar target of profit from my investment. I will set up my expert advisor like that. The cycle ends as soon as my target meets.

Next, I need to reassess my target, I need to reset my EA for the next target.

More to follow on Time Series Analysis (keep reading).

I would say the difference is if your equity increases your 1% stop loss distance will increase which is ofcource better than a fixed stop loss, Rekon-X EA uses trailing so it tries to move the stop loss in profit as soon as possible.

Both the statements are pretty close but not exact and yes, test result will differ.

But still the moving average metod can only ensure your winning 50% maximim and to make profit you need to use your intution and other signals coming from market.

Traditional Moving averages does not give you the visual of breaking down into the details what needs to reduce the error% from 50% to less than 5%. You need to have something better than MAs.

If you ask what is the difference between a Time series Analysis and moving average analysis. Then the major point would be which time frame? which cluster of time?

Here is a sample of NZDUSD trading on Friday based on time series analysis, can you spot at what time is the most profitable trading time for NU on friday. Let's apply the same on this friday and post your result.

Thank you for your view. I will be sharing more details as the day progresses. Let's concentrate on the contents and methods of the work.

It is people's choice to follow something or not to follow. All views are welcome.

I have removed the WhatsApp group name, in that I used to share free information, I think few vigilants here did n't like it. Hope it is alright now.

2. Supervised Learning: I will get back to explain the time series more, in the mean time let's understand the supervised Learning concept as this is the most essential part of RekonX method.

When we are talking about supervised Learning, it means we need to monitor the EA input, process and output.

For example, the EA first filters the most tradable time of a particular trading day.
This approach needs to be supervised and fine tune. Because if we continue use the same trading times over the period you will see slight changes on trading time market reaction behavior.

If, trading time 12:30 GMT is profitable for a quick buy of GBP last year, it may not be true for today trade.

Today may be selling GBP at 12:45 GMT is good for 30 minutes to 1 hour. And you get a signal to stay away from GBP for next 4 hours due to uncertain pattern reported by analysis.

You need to keep on doing the currency pair study, and supervise the EA's input eg buy GBP at 12:30 GMT on Thursday for 0.01 lot and sell GBP at 16:00 GMT on Thursday for 0.02 lot. A weekly review or refresh of Time series analysis is ideal.

You do it on Sunday before market open, may be 1 hour is enough to cover time series study of 10 different pairs.

Is this all about historical data study? Answer is no. Once your time series analysis is done you may run a linear regression to predict most profitable trading time of the instruments to reduce the error%.

Regression analysis is optional for now, we'll discuss later.

3. Selecting Currency pairs and engaging them to work for you during trading hours


Before reading further, let's do an exercise. When you trade which instruments you like to trade?
How long they take to meet the profit target?
How many hours on maximum you trade or your existing EAs trade?
Are all the trading hours you/ your EA' engaged to give you the targeted profit?

The ideal result should be -
To achieve 2% increase of today equity we have the following options -
A. We opt for quick turnaround - in that case we need to risk more say 10% with 80% probability. Question is how you ensure the probability, risk and reward.
B. We opt for a slow turn around a day long trading. Distribute the risk lower manner, target multiple time frames pick multiple trading options.
Eg. Instead trading 2 times GBP at 10% risk for 2% profit. We take 2% risk on 5 instruments say eurusd, usdchf, usdcad, usdjpy, eurjpy and target for 0.4% profit on each.
We check for >80% success probabilities on them and distribute the risk.

Equation will be ,

I trade 2 and I win 1 and lose 1
I trade 10 and I win 8 and lose 2

In first case, your equity can reduce 8%
In second case your equity will not reduce significantly like the above case and it will have higher chance to meet the profit target early and shut down for the day.

4. Every trading day is not the same

If you are using metatrader 5, just download the history in an Excel file. Create a pivot table and chart.

Create options to drill down the numbers by year, month, day, hours, and minutes.

You must have a list of high/ critical events and time when those were released or scheduled.

Mark the trading options and set the EA accordingly.

Risk reward is important, trailing stop is important because for critical news hours traditional approaches will not work.

You know it for sure, you need not to have 3 years observation for that. How flexible is your existing EA s to act on critical news days and hours?

So, in this EA it can be set to act on news trading as well.

5. Controlling trading lots, profit, stop loss

Based on your risk appetite it is better to have a fixed trading lot per day so your risk will be under control.

Then distribute the lots to execute >80% probable trades, trail stop, take profit , achieve the target and shutdown for the day. Resume on next trading day avoid market overlapping hours and closing hours.

When we have fixed target chasing and gets wrong entries we need stop and reassess the situation, the EA needs to understand this as well.

So whether we should have options like martingale in which in case of failure it will increase the stakes to risk more for achiving the target?

For now I have not included martingale feature, for future update I may keep it as an option. Also, I do not totally dismerit martingale when we are working with probability. But not including it now because I do not need equity absolute drawdown go below 50% at any point of time.

For now, it will be more conservative approach on meeting the target on weekly basis because even a daily target will be aggressive for any EA.

I have shared the WhatsApp link to all who are interested for RekonX EA testing. Let me give you some more details about the procedure.

As I mentioned earlier, before distributing the EA i need to make few changes on the demo product. Although it will have all functions and no limitation as such. It needs to be under super vised learning and u will be able to customize most of it's settings by your self only and deploy it to any vps or just enable auto trading at your local system.

It will come along with two csv files one for buy entries one for sell. The entries needs to be updated time to time. Weekly refresh is required also news event related update is required. You can use the same files coming along with the EA but the accuracy will be reduced every week. Just update them as per your time series study or keep watching this space for all weekly update on buy sell entries on major and minor instruments.

Example: EURUSD buy 16:00

This entry will trigger buystop and buylimit orders at 16:00 of your terminal time.

One more catch here about the trade times set. So based on your broker you may need to do a full time series analysis by yourself to review and finalise the trade timings.

I will be posting time series analysis from fxpro and icmarkets, I think the broker terminal time is same worldwide. Correct me if I am wrong.

On the trial phase I will try to occupy 8 hrs on a particular trading day, I will choose from major and minor pairs maximum of 8 pairs and deploy them for trading.

Risk reward ratio I'm going to compute assuming the bots are deployed on a usd 100 account. You need to understand it is pretty tough for bots to work on such conjusted and limited capital.

Since, it is going to have conservative approach so I am hoping it will still work fine.

Trading Lots will be 0.01. Autobots having this capability to fire 100s of orders within a second if your system is connected with a good internet bandwidth.

But since it is conservative, they will be firing limited orders only and trading fixed lots.

There are lot of work, I will try to make this available to you by Friday morning (IST). It may not work perfectly, don't be afraid and post your observations for betterment.