MACD / CCI and ADX in demo account: strange results

[Esthorus]

Hello everyone and sorry for my English, I modified an EA of MacD example by adding some things about Adx and CCI.(It is the first time I develop in MT4 and I am a rookie in trading

[Esthorus]

I upload the first capture;There was a problem in the first image.I don't know what the small green triangle means.I think it is specific of the original code.

[Esthorus]

Other information, the test has just finished: initial capital (January 01

[NorsesnNor74]

To see if there is any mistake, you should try it live.You will know it soon.I have seen many graphics like that;Don't get too emotions.

[Esthorus]

Thanks for your answer.I am not excited, only surprised.For me it is practically impossible to obtain benefits in all operations.Your answer confirms the strange of my graph.Next week I will try to do the test live if I have time.Do you know if history can generate this type of problem?

[crusgesr72]

Could you share the EA for others to try it?Thus you would receive many opinions in a short time and we could see if there is something strange.

[Esthorus]

Thanks for your answer, I will do it after another test.I saw that the expert was configured only for sales (Sell

[Esthorus]

Here are two new captures of the report.In the first, the great fall is due to the end of the history.Then, the EA stops the operation as it was, since the trend was bassist and my operation as well.Without that fall, the test is good, although with less gain due to the parity used.In the second capture: - Profitable short operations:> 95% - long profitable operations:> 80% The conditions to open a long position are different from those of a short position to reduce non -profitable operations, but these represent a large part of thetotal losses.If someone has ideas to improve this, I appreciate them.

[Esthorus]

Hello everyone, does anyone have any comments or suggestion?This morning, even in demo but in real time, I opened my EA in 14 pairs, only in short.M1 and proportion of 0.1.At the moment: - 4 open operations - 4 closed with benefits

[Esthorus]

Hello, I already discovered why I have such miraculous results.It turns out that the code does not close the operations correctly when it should, so wait for a long time until it becomes profitable.It depends on certain conditions, because sometimes it closes well and sometimes not.I still investigate.

[vuxxesvesS]

You may have discovered a failure in the exit logic of your EA.This type of errors is not uncommon, especially if it is your first time programming in MT4.Be sure to review the closing conditions well to see if there is any inconsistency in the code.

[1977Gistesva]

What you mention is quite common when the EA does not have a clear exit condition.Basically, what happens is that you are waiting for a reversal.This works in demo, but in real you could get stuck in a position for a long time.

[Costro]

For an EA, especially one so sensitive to time, you should verify that all output conditions are well configured.You could try to add a limit time condition if the Trade does not become profitable after some time.

[josesnto]

Those results are misleading if the EA does not close when it should.In real trading, every second account and the market will not always move in your favor.Give a review to your closures and make sure they don't depend only on luck.

[TaSaxcesrnu1]

From what you describe, it seems that the EA is "frozen" waiting for a reversal.This can work in a backtest, but if you test it live you will see that the results are very different.Try to configure more strict closure conditions.

[jiesnSius2323]

I think you are dealing with a "survival error", like many new traders.Basically, you leave open operations until they become profitable, which is only seen in historical graphics.In Real, those trades can give you surprises.

[ccesesrrCCaakneses]

If you are going to try live, be sure to use a small lot size.You better discover any potential error with a minimum investment, since demo results do not always reflect reality.

[TesaSesNodzz]

Try add a maximum time condition for the EA to close the position if it has not become profitable in a certain number of bars.This will make the results more realistic and that you do not depend on waiting for a reversal.

[inesuzg]

The small green triangle you mention could be indicating a specific condition of the original code, such as an operation opening signal or a data update.Check that part of the code and see if you can adjust it to make it less ambiguous.

[TesiSutesTesiSushi]

Did you try to test in other currency pairs?Sometimes an EA gives very good results in a specific pair due to the structure of the market, but in other peers it behaves differently.Test in different assets to see if the performance is maintained.

[choCuSagra]

Demo graphics usually give results that are not realistic.Try the EA in real conditions or in a micro account to obtain more reliable data.This way you will realize if it really is profitable or if the results are only seen in theory.

[DesnesySvesro]

The problem of the EAS is that they often have hidden failures that you do not see until the real -time tests.If you discover that it depends on waiting for operations to become profitable, that could be a problem in real market volatility.

[RvbonWost]

Keep in mind that, in Demo, historical data may differ slightly from those of the market in real time, which could be influencing your results.This is especially common with high frequency tests in low temporary frames.

[BresusTartoSes]

It seems that you have an "optimistic code", where EA simply is waiting.It is quite common and can be very misleading.I suggest you modify the code so that it has clear exits and does not stay "waiting for a miracle."

[rissaxnxx]

If you are going to make changes in the code, be sure to try in simulated environments.Sometimes a minimum change in output logic can make the EA become unstable, especially in periods of high volatility.

[axresyTkrros62]

I think the "miracle" you describe is simply the result that the EA does not close the losing operations.This works in demo, but in a real account it is a disaster waiting to happen.Give a review to that part of the code before taking it to Real.

[jarduSato]

Do not trust the evidence in demo.Often, historical data does not reflect the conditions of a real market, and the EAS can behave differently when the spreads and latencies change.Try a small account first.

[resNes1994birgaes]

Check the code you drive for closing conditions.It is common for rookies to leave those incomplete conditions and the EA simply leave open operations until, eventually, they become profitable.

[orukes2222]

A trick you can try is to add a "temporary stop" condition.For example, closes operations if they have not become profitable after 50 or 100 bars.This will cause EA to behave more as it would in Real.

[SuCakesdesbcn]

If you use M1 and you have such good results, it is almost certain that something is wrong with your output code.It is very rare for an EA to be profitable without loss in such a low temporal frame.Check the closures and make sure they are realistic.

[esnTknuagC84]

Try to make backtesting in higher time frames, such as H1 or H4.Sometimes E as show unrealistic results in very low frames, and when trying in Marcos Altos you can better identify errors.

[axresy4488]

If you are new in MT4, I suggest you review each function in the code.Sometimes there are logical errors that do not jump in demo, but when trying in Real, the EA fails because it does not consider all market conditions.

[JesSojesndraT]

That logic of "waiting" until an operation is profitable is not realistic.In the real market, changes can be fast and deep.The key is to have clear and, if they are not profitable, accept losses.

[Esthorus]

Thank you all for the suggestions.Now I am trying to add a time limit condition to close operations, as several recommended.Apparently it helps reduce those "miraculous" operations that seemed to be open indefinitely.I continue with the demo tests to refine details.

[esSCondrutes]

I am glad that you are implementing the limit time;It is key in any EA.If you can, try to vary the closing times and compare the results.You may discover that a 50 candles limit works better than one of 100, or vice versa.

[nokonsc]

Let me be direct: an EA that does not know when to close is like a drifting ship.You can have good results in demo because, eventually, all trades win over time, but in real, you could get caught in huge losses.

[Gattungortt]

I would tell you to try a cents or micro account.It is the cheapest and safest way to verify if the EA really works in real conditions.Do not trust demo too much;As soon as you enter real, everything can change.

[Torwzskcnes]

If you do well with the time condition, consider also try with fixed gain/loss limits.Sometimes, when EA only depend on time to close, they end up trapped in long trends that do not anticipate.

[CaCadobosor]

This type of EA usually works very well in ideal conditions, but the reality is another.When you enter Real, you will see how losses accumulate if the market does not go in the desired direction.

[RaboryTungi]

I think your EA depends too much on luck to be reliable in real trading.You should try a code that limits short -term losses instead of waiting for a reversal.This will prevent your account from becoming unsustainable.

[esSborTkbaSakesy]

Be sure to check the Slippage.In demo, you will not notice it so much, but in a real account the slip can greatly impact the result.Your EA could be perfect in demo and fail in real just because of that.

[UnnosCux]

How ironic that you are so surprised with the results, when it is clear that the code has failures.Waiting for all trailes to be winners is a bit naive, don't you think?Better works to make the EA more "realistic."

[JiesnxuTk]

You may want to consider a dynamic stop parameter, which adjusts the closure of the operation depending on the market.That could help you not depend only on time or "return the trend."

[ugartttuzes]

If you see that the operations are still open too long, check if you have any condition in the code that is blocking the closures.Sometimes, by adding functions, we forget those details that end up affecting the final result.

[Barr?s6]

If you do not set the closures well, your account could be trapped in a lost operation for weeks.Well -designed EAS have clear exits to avoid this.Keep testing in demo until operations close without needing to wait "miracles."

[Tuchakesaxy]

Some traders recommend trying an EA in high volatility conditions to see if it resists.Have you tried?If your code does not handle closures right now, it is a sign that you can fail in real.

[sorgaSaSa98]

If you want reliable results, I suggest you try the EA in real conditions during news events.There you will see if you are really prepared for challenging market conditions or if it only works in an "ideal" environment.

[jesCakora24]

Although they work well in demo, it does not mean that it is profitable in real.Remember that demure simulation in demo does not always represent the reality of the market, especially when it comes to low time frames.

[Siuyu849]

You can try configuring percentage risk limits instead of fixed times.Sometimes, this makes the EA more flexible to changing market conditions, without waiting for constant reversion.

[Soesrsucakstriccuakos]

The EAS are very rafficing in demo, and what you see is not always reality.In a real account, you could see slippage and latencies that totally change the results.The only way of knowing is testing in real conditions, even with little capital.

[SiTuak79]

It seems to me that you are falling into the typical mistake of trusting perfect data.MT4 in demo is very optimistic.If you really care about the effectiveness of the EA, try a cents account to see reality.

[jriNSa]

Since your EA only opens operations in one direction (short

[noroesNugw]

It is rare that no trace has been loser.If the EA is only waiting for a reversal, that is not a winning system.It works in Backtests, but in Real, you could lose fast money if the trend is still against.

[Dog19]

Ideally, you add some capital protection measure, such as a dynamic trailing or stop loss.With that, even if the position goes wrong, the loss can be limited and avoid waiting for a miraculous reversal.

[Sttesdes14071]

Sometimes backtests don't reveal everything.In the real market, the Spreads change constantly and can affect entrances and outputs.Be sure to try the EA in a cents account to avoid unpleasant surprises.

[cSesidesy2424]

I see that you are really trying to improve the EA, and that is good.Test with high volatility events and observe how it behaves.An effective EA must be able to handle both stable and chaotic times.

[Giuxxoraxun]

The "miracle" you mention is not such if the EA has no clear exits.Imagine that you are caught in an operation for weeks.In Real, that would affect your capital and would leave you without maneuver.

[axnTkb01]

It seems to me that your EA is very optimized for one pair.I recommend you try it in other assets to see if logic is really solid or if you only adapt well to Eurusd in backtesting.

[esesrakesSotu]

If you have added the CCI and ADX, why not use those indicators also to define closures?You could try that the EA closes when, for example, the ADX indicates a fall in the trend.

[NonrusoSbaba]

The function of "waiting for it to become profitable" sounds unrealistic.Although you are working in demo, you should add rules to get the market if the operation goes wrong.Patience is not always a good ally in trading.

[TesrsoruTk027]

You should test with data from different years, because sometimes an EA works well in a specific period, but when you apply it to another era, it ceases to be profitable.It may only be optimized for a market type.

[DesnuoSswca]

Have you tried to use your EA in different time frames?Sometimes the results change dramatically if it moves from M1 to M5, for example.It is good to try different scenarios to see how it behaves.

[esdresyncojutes]

It is rare for anyone to mention this, but make sure your historical data is of good quality.Backtest results can be misleading if the data is not precise or if there are holes in the information.

[rescesTiso]

To ensure that the EA is solid, it is proven to use it in lateral and strong trend conditions.The best EAS usually work on both scenarios.If it is only well in a trend, it could be optimized for a single type of market.

[JesvuNaktesSbes]

The truth, it seems that you trust Backtest too much.I have seen many EAS that are great in demo, but when the real moment comes, they lose.Do not trust you so much until you try it with real money.

[Toroses10R]

If you are not losing in any operation, there is something strange in the code.Unless you have a system that is 100% precise, that lack of losses only indicates that you are waiting too much in each operation.

[JtthDodo82]

Having an EA that does not close positions can easily seem good idea in backtest, but in real trading that is a recipe for disaster.You could get caught in huge losses and unable to close on time.

[AxyoSo]

Did you consider optimization of your EA parameters?Sometimes, changing the value of ADX or CCI a little makes the EA less dependent on time and more adaptable to different market conditions.

[EsCahunes]

I think you are getting carried away by the results in demo.The graphics always look good there, but in real history is another.Without clear output control, you risk losses.

[EsSNak3kesn]

If you really want to prove its effectiveness, use a real account with a small capital.Thus you will know if the strategy works or if it is only a "perfect in demo" that does not adapt to the real market.

[Trikku]

Try to add an output condition based on ADX.For example, if the ADX lowers a certain level, the EA should close the operation.This could give you more reliable exits without depending on the market recovering.

[sesnddd]

Although you try it in demo with good condition, solidity is required in real trading.If your EA is inconsistent with closures, you probably have problems when the market becomes unpredictable.

[RaCy]

I recommend that, in addition to ADX and CCI, you include some type of trailing stop to protect profits.Thus you prevent an operation that goes well to lose value just because the EA does not close in time.

[skesTaT]

Have you tried to add a mobile average to confirm the outputs?You could configure the EA to close if the price crosses a mobile mean in the opposite direction.It is a simple but sometimes effective method to avoid getting caught.

[CesrSesEssonsua24]

If you really want to see how it behaves in Real, try a penny account.Although demo results are good, a real account will tell you if it is a solid EA or only an illusion of the backtest.

[GarkesoN]

The lack of clear closures in your EA is worrying.In demo, of course the results are good because you have no real risk.But when you operate in real, you could face huge losses.

[jttoSakaa]

To improve EA stability, try to optimize each indicator based on historical data and adjust its values.Sometimes just changing the value of a parameter improves closures without compromising profitability.

[CEsRSTTcesrSates]

If you are going to keep it alone, make sure the closures are precise, because if the market enters an upward trend, that EA will be trapped.