Second day of trading Forex - I love stop losses!

[cogs]

Nickf,

Did you run your scripts through the last couple days? The best days for testing in my opinion.

I had a few EA's running with a loose and overleveraged setting on EURUSD so expected some damage, but otherwise they survived well. I see some brokers even stopped new demo accounts from being opened, whether it was too much server load or just a great time for extreme testing.

 

[NickF]

Nickf,

Did you run your scripts through the last couple days? The best days for testing in my opinion.

I had a few EA's running with a loose and overleveraged setting on EURUSD so expected some damage, but otherwise they survived well. I see some brokers even stopped new demo accounts from being opened, whether it was too much server load or just a great time for extreme testing.

Hi Cogs,
I actually run three scripts on three demo accounts. Two of them overall made some money, although they've been losing a lot as well - I didn't even realize they were ahead. I will show the transaction curve for the winning ones. All scripts are working on about eight instruments each on every account. Unfortunately I didn't have time to optimize the parameters for every instrument and in fact I've done some partial optimization for a few instruments and applied it to all the others.

I am not particularly happy with either script, but they've done better than me (on Friday I lost $440 and so far I recovered it all today and I am a bit ahead at the moment). I also added another $800 to my account, now the total deposit is $3800 and I have equity of $2517 (could be very different story tomorrow, as it varies quite widely).

But I think there is room for improvement in my scripts - I will have to check if the moment they entered a trade was any good and if it was, I will see if I can improve the exit. One script is definitely making far too many trades, while the other one, not enough... I was hoping if I set the parameters to find trades quite rarely, the success rate would be high (but it didn't seem to be that way). In one script, I am using a trailing stop loss and I am not very happy about it. Maybe I need to give it more room to breath, I think by setting the SL too close to price, I kill it too soon.

The first screenshot I show may not be too informative - since I work on scripts and coefficients quite often, the result may not indicate a long term trend. This screenshot shows all trades from 03.06.2016 - 27.06.2016 and in this period I made some changes to both the script and coefficients.



The second script runs too many trades and it has a poor success rate. I selected only the trades between 24.06.2016 and 27.06.2016, because this is my main development computer and I lose a lot of money during testing (especially when I play with 1 lot).



Anyway, after checking these two graphs, I see that I should not complain too much - it could be better, but it's not too bad either.

Nick

 

[cogs]

Looks like you may be on the way.

After seeing many manual pros and algos running it appears that a win percentage up to approx 66-67% seems to be a ceiling, and about as best as most can reach, there may be higher than that, but it does not seem to be sustained. If there are, I am sure there are but they would be big dollar setups.

Your max drawdowns look a little high, having said that anyone who trades FX (if you do), needs to develop a stomach for drawdowns as we have DD right from when we open a position. Big players also know that retail has less tollerance for DD.

Maybe I need to give it more room to breath, I think by setting the SL too close to price, I kill it too soon.

Yeah, trailing stops, I went away from the conventional, as they are very inefficient and simply ratchet in to market price. Using conventional trailing stops I found I needed approx 60pips before starting to trail simply to reduce the ratcheting affect. I have my own opinion when using stops, and quite simply whenever I found myself playing around with stops I found I should be looking closer at entry instead. Stops work great on demo and strategy test, say no more.

I am not sure what type of entry you use, but opening at market price seems to be a common mistake most make, as we try to find a perfect entry, which simply does not exist.

Another good analysis tool for taking a close look at results is MT4i desktop for when you use MT4 that is. If you can stomach the DD your chart results look quite good, don't they, and extrapolated out over a longer time frame could be very impressive.

Keep up the great work!

 

[NickF]

Looks like you may be on the way.

After seeing many manual pros and algos running it appears that a win percentage up to approx 66-67% seems to be a ceiling, and about as best as most can reach, there may be higher than that, but it does not seem to be sustained. If there are, I am sure there are but they would be big dollar setups.

Your max drawdowns look a little high, having said that anyone who trades FX (if you do), needs to develop a stomach for drawdowns as we have DD right from when we open a position. Big players also know that retail has less tollerance for DD.

Yeah, trailing stops, I went away from the conventional, as they are very inefficient and simply ratchet in to market price. Using conventional trailing stops I found I needed approx 60pips before starting to trail simply to reduce the ratcheting affect. I have my own opinion when using stops, and quite simply whenever I found myself playing around with stops I found I should be looking closer at entry instead. Stops work great on demo and strategy test, say no more.

I am not sure what type of entry you use, but opening at market price seems to be a common mistake most make, as we try to find a perfect entry, which simply does not exist.

Another good analysis tool for taking a close look at results is MT4i desktop for when you use MT4 that is. If you can stomach the DD your chart results look quite good, don't they, and extrapolated out over a longer time frame could be very impressive.

Keep up the great work!

Thanks Cogs,

I learnt that a high success ratio is not necessarily the recipe for success, so I rather aim for consistency and lower DD.

Coincidentally, I am also about to consider a better entry point. I have in mind to put half of money when I enter the trade, then use another half at a better price. If it happens the trade goes the right way immediately (which is rarely the case), I profit of half the funds. If it goes the other way, I buy some more - but only once (or maybe twice?).

The drawdown is hard to digest, but long term, it looks like the more DD one is willing to accept, more money is to be made - of course, provided the strategy has an edge. I found out that when I get itchy to close a losing position, that would be the best moment to buy some more - too many times I closed a trade near the peak of a reversal. But perhaps despite this "bad luck", being cautious maybe saved my skin a few times. So far, I'm grateful I survived the first 15 months and now I appreciate a bit more the value of patience (even if I am not a patient person) and knowing it only takes hours or days to lose or win big time. Although, if it's not boring, it means it's too risky.

Nick

 

[NickF]

Today could have been my best day in Forex, ever. All my trades were on positive and I caught a few big drops of the GBP and I ended up with a net profit of slightly over $500. I got so excited about the huge profit made in only one day, that I lost contact with reality.

I was sure that the traders were panicking and selling at a loss and I could not see an end to the drop. I was waiting to see a profit of $600, then $700 and so on. I didn't care when the price started going backwards and the profit diminishing. Every few minutes the price seemed to start dropping again, so I was hoping again for reaching again the $500. I was not too concerned when the price reached $450. Not too concerned about $400. When it got to $350, I thought - I didn't sell at $450 or $400, why should I sell now at $350.

To cut a long story short, I closed all my transactions for a profit of $120. This should still be very pleasing, if I ignore the fact I somehow managed to lose all the rest to $500. If somebody told me this morning that I will reach $500 and then I will lose almost all of it, I would have told they're dreaming

So, at the moment I am more unhappy for making $120 today than when I previously lost $440. My account is at the moment $2881 (while earlier in the day it reached $3200)...

I also added two screenshots just to show how badly I am timing my exits. No matter if I wait an hour, or four, I seem to pick up one of the worst moments...





On the good side, one of my scripts (the more primitive one), seems to perform quite well.

It started with $5000 on 21.05.2016. Unfortunately I can't run it on my current FXCM account, which only works with Trading Station. I created another proper account, but I don't have funds in it.



Nick

 

[Jason Rogers]

Today could have been my best day in Forex, ever. All my trades were on positive and I caught a few big drops of the GBP and I ended up with a net profit of slightly over $500. I got so excited about the huge profit made in only one day, that I lost contact with reality.

I was sure that the traders were panicking and selling at a loss and I could not see an end to the drop. I was waiting to see a profit of $600, then $700 and so on. I didn't care when the price started going backwards and the profit diminishing. Every few minutes the price seemed to start dropping again, so I was hoping again for reaching again the $500. I was not too concerned when the price reached $450. Not too concerned about $400. When it got to $350, I thought - I didn't sell at $450 or $400, why should I sell now at $350.

To cut a long story short, I closed all my transactions for a profit of $120. This should still be very pleasing, if I ignore the fact I somehow managed to lose all the rest to $500. If somebody told me this morning that I will reach $500 and then I will lose almost all of it, I would have told they're dreaming

So, at the moment I am more unhappy for making $120 today than when I previously lost $440. My account is at the moment $2881 (while earlier in the day it reached $3200)...

Don't beat yourself up about not catching the entire move, Nick. If you buy at the exact bottom or sell at the exact top, that's luck not skill. Congrats on a good day of trading!

That said, what made you decide to close with the $120 profit? Also, earlier in the day, what made you think the price would continue moving in your favor when you were up $500?

 

[cogs]

That said, what made you decide to close with the $120 profit? Also, earlier in the day, what made you think the price would continue moving in your favor when you were up $500?

FXCM Rep.

Do you trade, if so can you show some of your decisions and live trades?

Or do you simply carry out research via forums for FXCM?

 

[NickF]

Don't beat yourself up about not catching the entire move, Nick. If you buy at the exact bottom or sell at the exact top, that's luck not skill. Congrats on a good day of trading!

That said, what made you decide to close with the $120 profit? Also, earlier in the day, what made you think the price would continue moving in your favor when you were up $500?

Hi Jason,

I totally agree with you about catching the bottom or the top as pure luck/bad luck. I sold at a $120 profit only because I did not have the mental spirit to see a very good day turn into a losing day. I did not know if it will continue to go against me into negative territory. And I thought that a $120 is still excellent profit for me and take it, while being wiser next time. It turns out I made a mistake - I should have kept it, since I still expect the GBP to keep dropping.

When the profit got to $500, I had the strong feeling to sell all what I have and then wait for prices to bounce back where I could buy again. But I've been somewhat psychologically affected in the past and now I am more afraid of losing a greater opportunity than losing money. If that $500 would have potentially transformed in $700 in a matter of minutes, I was afraid I would regret for selling too early. But of course, ultimately it was my greed that cost me losing the money I had in hand.

When I considered selling for close to $500 profit, I hesitated because I would have had to close each instrument separately and that takes some time for me, on the mobile. I wish the mobile application of Trading Station would offer a way to close all trades at once - I don't think there is such a feature. I am a bit slow in trading on mobile phone. However, that is really no excuse - I had many hours of prices going up and down and I noticed the higher highs and higher lows over the entire period. Well, I have to make peace with myself - it was easy come, easy go money.

So far I still have the best month ever. Because now I added funds in total of $3800 and I have at the moment almost $3000, I am quite comfortable with higher risk and don't panic as much when prices go against me - sometimes I add a bit to my positions - at this moment I am trading in total $15000 in three pairs. Yesterday I noticed a maximum profit of $80, which I hesitated taking it - it bounced back very quickly so I didn't have time to meditate too much. Because I believe the GBP will continue falling for weeks or months to come, and I noticed that many times when I target a short term profit, even if I make a good profit every day, it is nowhere as good as if I kept the pairs over the entire week. There is a huge potential for somebody to earn 5 times more by doing multiple trades per day rather than holding for a longer period, but that is not me. For example, if I hadn't sold my pairs for $120 profit, I would be in a much better position now. But I still managed to make a bit of profit the last day, yesterday I was up $80 and this morning I was left with just $13. Hopefully on Monday I will get back toward $80, or maybe even better.

It is easy to develop a false confidence in myself when things are going well for me. At the end of 2016 I will be able to draw a more accurate conclusion if my trading has improved compared to 2015. So far this year I had four negative months and three positive. I am 5% in profit, but we never know what tomorrow brings.

I see the game of trading similar with the training of athletes. It takes a huge amount of training to improve very slightly. And mental strength is extremely important. Yesterday I was watching a game of tennis, Roger Federer vs Milos Raonic. In the beginning, Federer was losing so badly, my wife was wondering why did he consider playing at Wimbledon anymore. But he had the strength of character to keep fighting and he was able to recover the tremendous disadvantage he had. He still lost the game, but in a very decent manner. Probably such attributes are beneficial for traders as well.

Nick

 

[NickF]

My EA that was apparently working well started to lose money quite a lot. I think it's my fault on one side, because I tweaked the parameters, to make it do trades more often. On the other hand, I think it performs better when there is high volatility, with major changes in price and high volume and not so well for average volatility. I'll try to improve it, by making it to work in reverse when the volatility is not high enough. I made my life hard by testing the strategy on multiple currencies (to increase the number of trades) and now I have to check if my simulation with past data shows the same results as I obtained during live trading...

Nick

 

[cogs]

So you are discovering your EA is only profitable in certain market conditions?

I think the next step is adding filters until it hardly trades at all, (he sais sarcastically). Without sounding like a smarta$$ or giving too much away, you may need to add flexibility, or smarten into your entry method.

A simple method of test I use, is an EA must be able to be profitable on all pairs some extent, that have reasonable spread, if it is only profitable on one pair, it will fail eventually, for sure. The rationale, every pair behaves like the other (more or less agressive) at some stage, so eventually the same market entry or exit conditions will be encountered.

Without stating the obvious, it does look like your stops are being hit in your losing streak? Your profit curve actually looks like an 1H or 4H line chart.

Sounds like you are keen, so you will sort it out. If you are keen and enjoy coding it really does make a difference to your end results.

 

[NickF]

So you are discovering your EA is only profitable in certain market conditions?

I think the next step is adding filters until it hardly trades at all, (he sais sarcastically). Without sounding like a smarta$$ or giving too much away, you may need to add flexibility, or smarten into your entry method.

A simple method of test I use, is an EA must be able to be profitable on all pairs some extent, that have reasonable spread, if it is only profitable on one pair, it will fail eventually, for sure. The rationale, every pair behaves like the other (more or less agressive) at some stage, so eventually the same market entry or exit conditions will be encountered.

Without stating the obvious, it does look like your stops are being hit in your losing streak? Your profit curve actually looks like an 1H or 4H line chart.

Sounds like you are keen, so you will sort it out. If you are keen and enjoy coding it really does make a difference to your end results.

Hi Cogs,

Thanks for the tips. The good thing is that I put my EA to send me email for every transaction it opens and it keeps track of EA version (I change the version every time I modify something), all values of parameters I had at that stage, account balance, profit, open trades, etc. Probably even better would be to write entries into a log file, rather than chase the emails.

The graph unfortunately does not show the time on X-axis. The profitable trades happened rarely, while the bad ones happened much more often.

I discovered that the losing streak only appeared since July 6, so in only last three days it did all those bad trades. As I mentioned before, it takes a long time to get optimal parameters for each currency (over 24 hours per instrument). And because the trades were happening sometimes as rarely as one per week, I was not sure if the EA still operated. Now I fixed that - during the day, the EA sends emails to me every hour and it informs about open trades and balance.

Besides that, I am still not sure why the strategy tester gives different results compared to actual trading. I run backtesting on all currencies since July 06 and I could not come up with a total loss as big as it shows in the graph. And I am running backtesting on every tick.

Anyway, chasing every email since Jun 3, I noticed that I made changes to parameters twice. On Jun 23, I set all instruments with identical parameters. On Jun 24 I had big change in balance - from $6011, it dropped to $4971 and later the same day to $6376.

On Jun 30, I was not happy because trades were happening too rarely, so I pushed the EA to perform trades more often by relaxing the entry rules. I also skewed the Stop Loss and Take Profit (I chased more profit and less loss). I wanted more excitement and I got it - but not in the expected direction

Well, that should teach me a lesson - to use optimized parameters, rather than guessing the parameters and be more patient. Good trades don't come too often.

Compared to your method, which seems to perform on every instrument with the same set of parameters, I am trying to use optimized parameters for each particular instrument. I will revert to optimized parameters and wait another month and see if the balance improves. Perhaps I should look if I find an "universal" set of parameters, that gives reasonable results for all instruments. I'll keep that in mind.

Looking back at what I've done, it's like I try to build a violin. But not only I don't know how to build the violin, I also don't know how to play it - so even if the violin would sound good in expert hands, to me it doesn't sound good, so I tweak it

What I could do now is to backtest with the original set of parameters for each instrument and see if the result would be better. If only the backtesting would give the same results as the real trading...

Nick

 

[cogs]

If only the backtesting would give the same results as the real trading...

From my experience this is the key, so your tweaks mean something.

Which results don't match, your entries or exits?

With an engineering background I like to analyse but also consider the 'human factor', I found only 2 conclusive and accurate ways to know if my changes were correct or not, data collection/analysis and back testing that matches live results. Live exits will hardly ever match back testing results, because once your position is in the market your position forms part of liquidity, so is fair game for others searching for liquidity. Because MT4 has friendly export methods, I used to run many spread sheets to analyse results, chart it up and .

Looking back at what I've done, it's like I try to build a violin. But not only I don't know how to build the violin, I also don't know how to play it - so even if the violin would sound good in expert hands, to me it doesn't sound good, so I tweak it

Sounds like a 'Quants' approach to trading.

 

[NickF]

From my experience this is the key, so your tweaks mean something.

Sounds like a 'Quants' approach to trading.

Good one Cogs

I will investigate which results don't match (entries or exits). But considering I am doing paper trading (therefore I don't alter the course of trading with my own trades), I don't think I should see a difference in either entries or exits.

Meantime I discovered something interesting. During the week, I can do my optimizations on FCXM MT4 Platform. But on Saturday, FCXM has the servers closed and I can't use it. Because for me the weekend is the time when I have most time to play with things, I opened a demo account with NoorCM and I run optimizations using their MT4 and their data. Today I had the idea to compare results, using the same strategy and same settings. When doing simulations for 2016, the results are very different - FXCM shows a profit of $3128 and NoorCM (where I've done the optimization) shows a profit of $7569. If I optimize on FXCM and then test on NoorCM, I should see a higher profit on FXCM. In my EA, I am using both price and volume and as far as I know, every broker has its own interpretation of volume (perhaps it is actually the number of trades counted by that broker during the bar and not related to the value of the trades - if that's the case, then 10 trades of $1000 would appear as a higher volume than 1 trade of $1000000).

Until now it was not important to me on what platform I was doing the optimization (because I did not expect significant differences) and it was clearly a mistake.

It would be interesting to check with a script based only on price - the results should also be different (every broker would be able to close the deals with different partners), but the question is how different would the results be. I don't expect the prices to be extremely different. A smart algorithm on the broker side would be able to chose the best partner for trading the currency. Speed is also very important, because a slower algorithm would miss the best opportunities. But this is way out of my league, I'm just speculating.

Sometime ago I came across a page that shows the profitability of Forex traders, based on broker (from 2013): http://piphut.com/2013/11/what-percent-of-forex-traders-make-money-is-trading-forex-profitable/

It shows that the most profitable traders were at the time with Interactive Brokers, while other brokers are less profitable. For example, FXCM had about the same number of accounts as Interactive Brokers, but the profitability was of 28% vs 44%. It could be that because IB has a much more complex trading platform, traders with more experience were using it. I've been using Trading Station from FXCM since I started and I enjoy the platform. It is well designed, reliable, has nice features and it is easy to learn and use.

It would be interesting to find if in more recent years the profitability by brokers has changed.

Nick

 

[NickF]

I found another profitability report (from Q4 2015) in here:

http://www.financemagnates.com/fore...-profitability-report-shows-improved-trading/

Interactive Brokers are still on first position and the figures have improved only slightly.

 

[Jason Rogers]

FXCM Rep.

Do you trade, if so can you show some of your decisions and live trades?

Or do you simply carry out research via forums for FXCM?

Hi Cogs,

While my role on this forum is to discuss forex with other traders, particularly FXCM clients, and answer questions about our platforms and services, I also trade personally. My main strategies use the Speculative Sentiment Index (SSI) as contrarian indicator, and I post some of my trade ideas on other forums where I maintain a signals and strategies thread. I'm not sure if I can post a link to those here, but please send me a PM if you have more questions.

 

[Jason Rogers]

Hi Jason,

I totally agree with you about catching the bottom or the top as pure luck/bad luck. I sold at a $120 profit only because I did not have the mental spirit to see a very good day turn into a losing day. I did not know if it will continue to go against me into negative territory. And I thought that a $120 is still excellent profit for me and take it, while being wiser next time. It turns out I made a mistake - I should have kept it, since I still expect the GBP to keep dropping.

When the profit got to $500, I had the strong feeling to sell all what I have and then wait for prices to bounce back where I could buy again. But I've been somewhat psychologically affected in the past and now I am more afraid of losing a greater opportunity than losing money. If that $500 would have potentially transformed in $700 in a matter of minutes, I was afraid I would regret for selling too early. But of course, ultimately it was my greed that cost me losing the money I had in hand.

Hi Nick,

I got the impression from you that might be the case, which is why I asked. It's important to have an exit strategy before you even open your trades. This might be based on a price target, or signals you might receive from indicators on your charts. As I mentioned to Cogs, I use SSI, as one of my primary trading tools, so I get my closing signals based on when the market sentiment has turned on a particular currency pair.

When I considered selling for close to $500 profit, I hesitated because I would have had to close each instrument separately and that takes some time for me, on the mobile. I wish the mobile application of Trading Station would offer a way to close all trades at once - I don't think there is such a feature. I am a bit slow in trading on mobile phone. However, that is really no excuse - I had many hours of prices going up and down and I noticed the higher highs and higher lows over the entire period. Well, I have to make peace with myself - it was easy come, easy go money.

I'm not sure how many trades you had open, but you can call the FXCM trading desk anytime 24 hours a day, and ask them to close any or all of your open trades over the phone. Also, I don't think there's any easy money in forex. Your trades that go more smoothly than the rest are the fruits of hard work you've already sewn and are also meant to carry you through the barren periods that will inevitably come.

 

[Jason Rogers]

It would be interesting to find if in more recent years the profitability by brokers has changed.

I found another profitability report (from Q4 2015) in here

Hi Nick,

Unfortunately, those reports don't take into account the difference in account opening minimums between brokers. Therefore, what they end up showing is the correlation between account balance and profitability which is also confirmed by our research.

It's important to note that FXCM offers services to traders at all levels, including Mini accounts ($50 minimum), Standard accounts (2k minimum), Active Trader accounts (25k minimum) and multi-million dollar institutional (FXCM Pro) accounts. By contrast, the other broker you mentioned has a minimum account opening balance of $10,000.

Below are stats from a DailyFX study on trader profitability in relation to account balance based on thousands of FXCM accounts. It showed that FXCM clients with an account balance of at least $10,000 (which is the minimum for the other broker you mentioned) had a profitability percentage of 43%.

​

 

[NickF]

Hi Nick,

Unfortunately, those reports don't take into account the difference in account opening minimums between brokers. Therefore, what they end up showing is the correlation between account balance and profitability which is also confirmed by our research.

It's important to note that FXCM offers services to traders at all levels, including Mini accounts ($50 minimum), Standard accounts (2k minimum), Active Trader accounts (25k minimum) and multi-million dollar institutional (FXCM Pro) accounts. By contrast, the other broker you mentioned has a minimum account opening balance of $10,000.

Below are stats from a DailyFX study on trader profitability in relation to account balance based on thousands of FXCM accounts. It showed that FXCM clients with an account balance of at least $10,000 (which is the minimum for the other broker you mentioned) had a profitability percentage of 43%.

​

Hi Jason,

Thanks for your answer. It is obvious that many $50 accounts would have a very short life compared to a larger account, so the research I read elsewhere does not compare apples with apples. I am satisfied with your explanation.

Nick

 

[NickF]

Oops, I did it again!

In my last trading day I lost over $800. I got relaxed, confident because of a pretty healthy balance, got a false sense of security because, despite reaching a total of $25000 in a very volatile market, my usable margin was 95%. Kept hoping my imaginary resistance levels will bounce the price back.

I didn't look that the GBPJPY has fallen sharply for six weeks in a row. I didn't take into consideration "what if the GBP will bounce back one or two weeks in a row"? In fact, I didn't even ask myself, what if GBP will increase strongly even ONE day only? I kept accumulating positions when the price "got better".

Now I am considering whether I should quit Forex for good. I am not concerned as much about my latest loss. I am more concerned about repeating the same mistakes again and again. Like not using risk management. Not being satisfied with a profit of 1% per day. Not concentrating more on learning than on trading. Being more interested in growing back my account instead of using good practices. Well, at least I got what I deserve

My account dropped from about $2950 to $2145 in one day.

 

[Jason Rogers]

I got relaxed, confident because of a pretty healthy balance, got a false sense of security because, despite reaching a total of $25000 in a very volatile market, my usable margin was 95%. Kept hoping my imaginary resistance levels will bounce the price back.

Sorry to hear that, Nick

Did you mean $2500? And how did you determine these resistance levels?

My account dropped from about $2950 to $2145 in one day.

My first thought reading this is that your leverage must have been extremely high. As mentioned in the article I linked to in my post yesterday, stats show that traders who use 10:1 leverage or less tend to have more success than those who use more than 10:1 leverage.