Parabolic sar expert advisor for mt4 (PLEASE) - page 2
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
OK
About my trading with PSAR. Here's my actual monthly statements... a good month and a bad month.
First the good month, then a bad month:
To say the least, after such great initial results (it went from $6000.00 to more than $10,000!), the end was depressing!
All trades were 1 lot. I had a standard account.
OK
About my trading with PSAR. Here's my actual monthly statements... a good month and a bad month.
First the good month, then a bad month:
To say the least, after such great initial results (it went from $6000.00 to more than $10,000!), the end was depressing!
All trades were 1 lot. I had a standard account.Interesting insight, thank you for sharing your true life experiences with PSAR.
d1 testing
hi Simaba
i appreatiate all your information on different threads very much. thanks also now about sar and td sequential very much. you tested only on D1 time frime oder? so the information is only about this timefrime?
regards
lodolYes,I tested only on D1 tf,for all majors,from 1992 for GBPUSD,USDCHF,USDJPY and from 2000 for EURUSD..frankly,what I expected to find was that either no single indicator worked alone by itself,or that if there was one,it was going to be the williams %r..what I found was totally different,as already explained.
I personally believe that markets are fractal,so,if it doesn`t work on D1,probably it is not going to work for lower tfs..and ,I don`t have clean data in metastock format for so long periods in lower tfs
When I do the psar EA if will be mainly for testing the assumption on h1,h4 tf and check if it works,with the above mentioned settings,for at least 3 of the 4 majors from 2001 to 2007
Regards
Insights
A few years ago, I wrote an EA using a fast and a slow PSAR on the GBP/USD 30M chart. I tried numerous entry methods but always exited using the slower PSAR.
For example, if I had a "BUY" position and the price collided with the rising fast PSAR I would switch to the slow PSAR and keep going in the trading position. If the fast PSAR value once again was lower than the price, I switch back to that for the stoploss. But eventually, the price would hit the slow PSAR as a stoploss and the trade would close. I would do the reverse for a "SELL" position.
This backtested very well, reaching over 1 million dollars when starting with $2000 in only 1 year. But I lost $3500.00 out of a $6000.00 investment when I traded with real money. This taught me an expensive lesson...what backtests well and looks good on charts may not pan out with real trading. Now I test my ideas with a live demo. It's much slower and doesn't build confidence very fast, but I hope that will imitate money-trading well enough to develop a working strategy.Hi jimven,
Thanks for your insights,they are very informative,and I would like to add something to them
1-The fact that a single indicator can be used both for entries and exits,with the same settings,profitably,for at least 3 of the 4 majorsdoesn`t mean it is not inadvertently overfitted..the 4 majors have practically the same dominant cycles,as shown by their internal correlation values
2-I have found complete systems that work unbelievably well for acouple of years and then self destroy in a few months..the concept behind the system was the actual overfit,for example,any trend following with pyramiding and a chandelier stop loss on d1 tf has for surely performed unbelievably well for the past 2 years on gbpjpy..but what about the ranging periods that inevitably are going to be there?
3-So,in my view,we have two different approaches for testing..the first one is to test for long periods(more than 5 years),where we can expose our system to a wide diversity of market conditions and ,if profitable,we will probably find that it makes money when trending/ranging and has contained losses during ranging/trending,so,we will learn to use lower position sizes than previously expected and look for an edge in the long run..
..the second one is to hyperoptimize a system ,every week, for the past 3 months and refit/hyperoptimize the parameters for the next week..sounds strange,and against all my beliefs on overoptimization,but I know 2 different traders that use this approach,both extremely profitable,one of them is profitable as in 3.12 Sharpe ratio,92.6% winning trades and 5.2 profit factor,though the ratio between averagewin and averageloss is only 0.42..I know because I have seen his audited results for the latest 12 months..and,no,I don`t know how does he achieves these results,nor what indicators does he use
So,if you want to test a system for real life trading,either test it on more than 5 years of data and check with other pairs ,same settings,to avoid overfitting..OR Hyperoptimize every week for the past 3 months,then you will achieve "adaptive overfitting"..In any case,just testing a system for 1 or 2 years and then going live is not an option I would recommend
Regards
i_3sar recommended use
Hi,
One way to use i_3sar is to "intelligently" filter the signals
If you take a look at the pic,you will see that the usdchf seems to be completing a downcycle ,at least in the H1 tf,so,the moment it breaks the farther away trendline it will probably be in a new confirmed h1 upcycle,then,and only then,I will take the i_3sar upsignal if there is any...
If you want to do an EA using this or other psar based indicators,I suggest that first you get used to its nuances ,then devise some kind of nuance-based filtering and you will find that it will help you inmensely in your trading
Regards
hi simba thx for answer. itried the indicator but didnt get the arrows on the same places like on your image. so for instance if i have 1H chart . ihave
PSAR with settings
0.032 for 1H
0.04 for H4
0.12 for Day(1440)
or did i get it wrong
thx for help
lodol
Settings
hi simba thx for answer. itried the indicator but didnt get the arrows on the same places like on your image. so for instance if i have 1H chart . ihave
PSAR with settings
0.032 for 1H
0.04 for H4
0.12 for Day(1440)
or did i get it wrong
thx for help
lodolThe settings,0.032.0.04,0.12 were for fibotrader D1 psar..forget about them with i-3sar..I just use either standard settings for h1,h4,d1,like in the pic..or 0.032/0.12..no need to overcomplicate,you are just loking for positive simultaneous momentum on 3 tfs..that`s all..sorry if my previous explanation mislead you
Regards
Simba:
I certainly appreciate your comments, especially those about optimization. When trading with a PSAR strategy, optimization was my constant challenge, but I eventually came to the conclusion that the PSAR CANNOT be optimized for any length of time because, for any specific 'step' and 'max' settings, it develops a certain 'pitch' than cannot adapt well to the changes in the frequency and range with which the prices migrate from their uptrend highs to their downtrend lows. These are somewhat chaotic, so the PSAR often rises or falls either too fast or too slowly. In order to be more efficient, it seems that an adaptive PSAR is needed that can change its rate of rise and fall to follow the strength (or slope) of the price changes. Often the PSAR move is far too shallow, and ends in a loss. At other times, it is too steep, and leaves plenty of profit 'on the table.'
I've experimented some with the idea of switching from one PSAR setting to another, dynamically, but haven't gotten anywhere with it. I've also experimented with some VERY slow PSAR settings that result in long runs in the hopes that the short, losing segments would be minimized. Nothing seems too satisfactory. Finally, I've experimented with such very slow PSAR settings to establish the overall trend and then, while they are active, trade with faster PSAR's 'inside' this longer one. Still, nothing that seems very workable.
As for Simba's methodology...I'm not familiar with it, although it seems that you are using MTF PSAR's to determine how to trade. I haven't worked much with that.
Good trading!
mtf psar
Simba:
I certainly appreciate your comments, especially those about optimization. When trading with a PSAR strategy, optimization was my constant challenge, but I eventually came to the conclusion that the PSAR CANNOT be optimized for any length of time because, for any specific 'step' and 'max' settings, it develops a certain 'pitch' than cannot adapt well to the changes in the frequency and range with which the prices migrate from their uptrend highs to their downtrend lows. These are somewhat chaotic, so the PSAR often rises or falls either too fast or too slowly. In order to be more efficient, it seems that an adaptive PSAR is needed that can change its rate of rise and fall to follow the strength (or slope) of the price changes. Often the PSAR move is far too shallow, and ends in a loss. At other times, it is too steep, and leaves plenty of profit 'on the table.'
I've experimented some with the idea of switching from one PSAR setting to another, dynamically, but haven't gotten anywhere with it. I've also experimented with some VERY slow PSAR settings that result in long runs in the hopes that the short, losing segments would be minimized. Nothing seems too satisfactory. Finally, I've experimented with such very slow PSAR settings to establish the overall trend and then, while they are active, trade with faster PSAR's 'inside' this longer one. Still, nothing that seems very workable.
As for Simba's methodology...I'm not familiar with it, although it seems that you are using MTF PSAR's to determine how to trade. I haven't worked much with that.
Good trading!Hi jimven ,
Using agreement of psar on 3 different tfs is just one of my entry methods,basically the first blue signal after several reds,or viceversa,is usually workable as an entry,but you need more to have a profitable system.
Regarding testing of psar,you are absolutely right..everybody ends up using 0.01,0.1 settings
..psar is like a wild horse,in trading that means it gives and takes away lots of profits and that is extremely frustrating..
Regards
thx now i got it.