- This topic has 315 replies, 40 voices, and was last updated 10 years, 7 months ago by
.
- Posts
This theory already checked years ago, and it’s Ok. The only thing to avoid is not to trade near current market extremes (i.e, SNB & the EURCHF at 1.20 level). Only trade within market “zone”. Another thing is to whithdraw profits at 20-50% levels, while the remaining account should continue with latest lot sizes. G.
Thanks a lot G. This will be a ‘killing’ field for the broker
A question: Is it right that a broker can ‘do something’ to our EA, after they know that our EA brings us good profit ? How to prevent this? ThanksThings that a broker can do: slight control of quote price Change slipage Change spread Delay quotes Requote G.
Thanks a lot G.
There is only one way of doing a backtest. First create a random history and let your EA ran against it. When it makes a profit, switch to the real history and let it run again. On the real history your EA has to be even better. And use as much history you can get.
Thanks muuh, a good idea.
Ah ok when the statistic is independent of the human and his trade it could work. I would go for the 99.9 percent after 2 days with a random entry. How are the statistics for 5 or 10 pips TP?
If you want to go for 99.9% after 2 days, it’s important that you reduce the TP % per trade to perhaps 0.20% per trade (this means reducing the lot size). Try it out with the EA on a demo account and see what you get
I plan on getting the statistics for tick data which I downloaded recently. I post the charts when I get the chance. What it will look like is an exponential decay where as you increase the TP the probability of success drops exponentially This is a good post. Thanks bro. Reducing the % per trade, means that we will put more trades in a day. But this is more secure, and train us not to be greedy … and the time factor (when) we open a trade is really matter, best on very very rare …
I got some interesting statistics for everyone. I tested this theory with tick data from January 1st 2012 to now for EURUSD. I checked the probability of success with a TP of 3pips (assuming 2pips for open/close spread). The probability of success came out to be 92.324% for a trade to end successfully within 24 hours! Total of 354,985 successful trades and 29,513 failed trades. What I did was, I opened a trade on every tick (short and long) and checked to see if it closed with 3pips in profit in both direction within 24 hours. So the theory is accurate, market is HIGHLY recurrent for a small TP! What this means is, even if we were to trade every tick blindly placing a trade for both directions WITH SL, we would still come out being a winner (assuming the markets and brokers were perfect with no slippage and news spikes, etc). Let’s do the math! 354985 successful trades => +354985pips (that’s 1pip profit each trade after we take out 2 pips for the round turn commission) 29513 failed trades => 29513*(3+2) = –147565pips (that’s with 3pip SL with 2pips for round turn commission) Total sum = +207,420pips of profit! In theory this is highly exploitable. If we can be good enough telling the trend, then the trades can be certain!
Hi Saver0! Love to see the outside the square thinking here.
I was with your thought experiment up until the 3 pip stop. You opened a BUY and a SELL and confirmed that there is a 90+% probability price will deviate in one direction 3 pips within the following 24 hours. Meaning this test didnt have a stop loss for the loosing side. So of the 29,513 failed trades, how many pips did each one lose?
Shouldnt the equation be 354985 winning pips less (29513 x average number of pips in drawdown after 24 hours)?
Initially you suggested using a 1 pip target with no stop loss. I can see the potential in your logic here (if losses can be managed).
However it seems counter intuitive for a 3 pip stop loss to be viable. Market noise is well beyond 3 pips.
Of the 1 pip winners how many pulled back >3 pips before hitting the target?
Furthermore, if you double your target to 2 pips (lets ignore spread etc for now) then you are doubling the amount of losses you can take. I would think that the probability of 2 pip targets would not drop very significantly from the 99+% you have calculated for 1 pip targets. I like your idea of tiny targets but there seems to be an imbalance when aiming for 1 pip.
What's it all about? It's all about money.
If I download tick data from other sources and import it to MT4. Are ticks would be virtual too? If yes, then MT4 is bad for backtesting, we need your own software to backtest tick EA’s.
One more thing. SaverO got about 400000 trades per 2 years. There are about 250 working days in year. 400000/250*2 = 800 ticks per day. It can’t be so little ticks per day.
EDIT:
One tick = 2 trades, so from 400000 trades must be 200000 ticks. 200000/500 = 400 ticks per day.
Thanks, IenDzi
I was with your thought experiment up until the 3 pip stop. You opened a BUY and a SELL and confirmed that there is a 90+% probability price will deviate in one direction 3 pips within the following 24 hours. Meaning this test didnt have a stop loss for the loosing side. So of the 29,513 failed trades, how many pips did each one lose?
Yes, I see what you are talking about and you are correct. Out of the successful trades, there would be a lot that would fail if they actually had a SL. So yes, the SL wouldn’t make sense. I will need to get proper stats with an actual SL to see what it comes out to be. If we take out the spread, it should in theory come out to be 50% I think since the entry is random for both directions. I will confirm it later. Thank you for correcting it!
One tick = 2 trades, so from 400000 trades must be 200000 ticks. 200000/500 = 400 ticks per day.
You are absolutely correct IenDzi! I feel so embarrassed for posting such inaccurate results. I looked at my script and I had loaded a sample.csv file with just a month of ticks when I had a 1.5GB file with millions of ticks. I’m running the stats for that file but it’s taking a really long time. I will let it run over night and post my results later. I’m hoping the stats will remain about the same though. 200k ticks vs 200 million ticks shouldn’t make much of a difference since its all probability but I will post soon my results
8 good trades as a test, but I need work on trend and entry.
Awesome results! Keep it up. Here are my trades from this week :)
Some of them I placed from my phone without care for the lot size since it’s just for testing (I was on holiday this week ). 22 total trades and if the lot size was accurate to get 0.4% from each trade, this would have been 109% gain for just 3 days of trading 
Focus, Patience, Determination & Order in chaos
This theory already checked years ago, and it’s Ok.
Thanks for the confirmation GG, means a lot
Focus, Patience, Determination & Order in chaos
You are absolutely correct IenDzi! I feel so embarrassed for posting such inaccurate results. I looked at my script and I had loaded a sample.csv file with just a month of ticks when I had a 1.5GB file with millions of ticks. I’m running the stats for that file but it’s taking a really long time. I will let it run over night and post my results later. I’m hoping the stats will remain about the same though. 200k ticks vs 200 million ticks shouldn’t make much of a difference since its all probability but I will post soon my results
Month stats sounds good for that ticks number. I think that the results will be the same for two years. It’s interesting, what would be win% with 2 or 3 pips. Do you test it with your own software or MetaTrader?
Today my demo trading:
20 random with my market understanding trades. Slowest trade under 9 minutes, fastest 6 seconds :D I think I was very lucky today.
IenDzi
SaverO are you setting your TP in the EA to equal 1 pip plus the average spread for that pair?
I have just been dragging my tp line after execution but on a phone that would be difficult.
SaverO are you setting your TP in the EA to equal 1 pip plus the average spread for that pair? I have just been dragging my tp line after execution but on a phone that would be difficult.
it’s 1 pip from open price:
When you sell it, TP is Bid – 1pip.
When you buy, TP is Ask + 1pip or Bid + spread + 1pip.
One tick = 2 trades, so from 400000 trades must be 200000 ticks. 200000/500 = 400 ticks per day.
You are absolutely correct IenDzi! I feel so embarrassed for posting such inaccurate results. I looked at my script and I had loaded a sample.csv file with just a month of ticks when I had a 1.5GB file with millions of ticks. I’m running the stats for that file but it’s taking a really long time. I will let it run over night and post my results later. I’m hoping the stats will remain about the same though. 200k ticks vs 200 million ticks shouldn’t make much of a difference since its all probability but I will post soon my resultsYou have got good result bro, really great. May i know how to create a statistic ? Especially if i want to know how much is the maximal draw down (in pips) for last 10.000 trades. Any clue or links to create some thing like that ? thanks
Edit: the purpose of this is to put a ‘safe’ SL, ex. our SL will be 1.5*maxDD.
One tick = 2 trades, so from 400000 trades must be 200000 ticks. 200000/500 = 400 ticks per day. You are absolutely correct IenDzi! I feel so embarrassed for posting such inaccurate results. I looked at my script and I had loaded a sample.csv file with just a month of ticks when I had a 1.5GB file with millions of ticks. I’m running the stats for that file but it’s taking a really long time. I will let it run over night and post my results later. I’m hoping the stats will remain about the same though. 200k ticks vs 200 million ticks shouldn’t make much of a difference since its all probability but I will post soon my results
You have got good result bro, really great. May i know how to create a statistic ? Especially if i want to know how much is the maximal draw down (in pips) for last 10.000 trades. Any clue or links to create some thing like that ? thanks Edit: the purpose of this is to put a ‘safe’ SL, ex. our SL will be 1.5*maxDD.
I think i got. It is simple actually … just look at High of each candle after signal candle (if price going up), compare it to previous highest, put the highest to buffer or variable. And compare the Low of each candle (if price going down). The difference to open price is the DD. Comparing each result on the whole trades will give me the answer.
I got some interesting statistics for everyone. I tested this theory with tick data from January 1st 2012 to now for EURUSD. I checked the probability of success with a TP of 3pips (assuming 2pips for open/close spread). The probability of success came out to be 92.324% for a trade to end successfully within 24 hours! Total of 354,985 successful trades and 29,513 failed trades. What I did was, I opened a trade on every tick (short and long) and checked to see if it closed with 3pips in profit in both direction within 24 hours. So the theory is accurate, market is HIGHLY recurrent for a small TP! What this means is, even if we were to trade every tick blindly placing a trade for both directions WITH SL, we would still come out being a winner (assuming the markets and brokers were perfect with no slippage and news spikes, etc). Let’s do the math! 354985 successful trades => +354985pips (that’s 1pip profit each trade after we take out 2 pips for the round turn commission) 29513 failed trades => 29513*(3+2) = –147565pips (that’s with 3pip SL with 2pips for round turn commission) Total sum = +207,420pips of profit! In theory this is highly exploitable. If we can be good enough telling the trend, then the trades can be certain!
So when you get to 3pips go to BE+2 with target price of 5 or 10pips. Why not go for a jackpot trade once you have nothing to loose
Thanks SaverO!!! Enjoying this topic!!!
Hi BillYon, nice to see you here …
So when you get to 3pips go to BE+2 with target price of 5 or 10pips. Why not go for a jackpot trade once you have nothing to loose
The win probability will drop much if our target are 5 or 10 pips. Trading with big lots will make our orders much risky (price can run in other direction before it reach 3 pips profit. 3 pips profit is equal to 4.5 pips if the spread is 1.5 pips). Trying not to be greedy with 1 pip TP can give us very big win probability (>= 99%). I think this is the reason Saver0 trade with very heavy lot.
Its been a long time i read something this interesting. Appreciated guys!
Talking abt the 1pip target, why not BE the trade after 1pip and let it run… till you have gotten your self too much profit or taken out? Will this affect the probability in any way?
regards
i was pointed to this thread by a trader friend, I too have been testing the theory & EA.
These are my days trades on a demo account, there is 1 which was an old order from before the test and 1 which I manually closed.
I experimented with position sizing and PT distance.
For me it only makes sense to have a PT greater than 2 pips, i will probably use 3 for the rest of my test.
3 pip targets
i’d also like to add , that having a PT greater than 3 pips doesn’t make sense to me when trading on the m1 chart. As if most traders put the SL 1-5 pips beyond the previous swing high low and we’re buying/selling into that SL run, we need to get out before all the stops are tripped and they reverse the direction.
3 pip targets
Hi all,
I’m back to my busy schedule. I managed to place 8 1pip trades for a total gain of 3.2% daily gain. In my books, anything around 1% gain is exceptional. I’m very happy and excited to see all the good trades you guys have been making. But it also makes me a bit nervous to see that we are making way too many trades and might be getting a bit over confident which could result in mistakes.
Please keep in mind the aim of this thread, that is to imprint in your brain how EASY it is to make 1% and that 1% is ALL YOU NEED!
Heck, if you could make 0.50% consistently, be happy, be very happy! If you can make 1%, be over joyed and go jump up and down with excitement because in 5yrs you will turn $1000 into $278 million. Please understand the magnitude of what 1% per day mean. Warren Buffet made billions making just 30% YEAR! 1% per day is over a 1,000% per year! That is an INSANE amount. No business nor index nor hedge fund ever makes this much.
The goal is to get just 1pip (or even less) because then you have probability on your side. You DO NOT want to rely on luck. If you rely on luck, luck will run out. Rely on hard math, probability.
I need to adjust my probability statements. I performed the experiment with 12million ticks and the probability came out to be 94.2% success rate for getting 3pips (with spread ~1pip TP). So this is the probability that we have trying to go for 1pip (with spread) if you were to take a random entry going in any direction without care of patterns or direction or trend. I’m certain that with following my rules outlined, that the probability of success for a trade within 48hrs is greater than 99.9% and most likely approaching 99.99%.
Don’t be greedy, go for 2-4 perfectly timed trades each day that are in line with the rules and you will have great success.
Green pips to all
Focus, Patience, Determination & Order in chaos
Hi all, I’m back to my busy schedule. I managed to place 8 1pip trades for a total gain of 3.2% daily gain. In my books, anything around 1% gain is exceptional. I’m very happy and excited to see all the good trades you guys have been making. But it also makes me a bit nervous to see that we are making way too many trades and might be getting a bit over confident which could result in mistakes. Please keep in mind the aim of this thread, that is to imprint in your brain how EASY it is to make 1% and that 1% is ALL YOU NEED! Heck, if you could make 0.50% consistently, be happy, be very happy! If you can make 1%, be over joyed and go jump up and down with excitement because in 5yrs you will turn $1000 into $278 million. Please understand the magnitude of what 1% per day mean. Warren Buffet made billions making just 30% YEAR! 1% per day is over a 1,000% per year! That is an INSANE amount. No business nor index nor hedge fund ever makes this much. The goal is to get just 1pip (or even less) because then you have probability on your side. You DO NOT want to rely on luck. If you rely on luck, luck will run out. Rely on hard math, probability. I need to adjust my probability statements. I performed the experiment with 12million ticks and the probability came out to be 94.2% success rate for getting 3pips (with spread ~1pip TP). So this is the probability that we have trying to go for 1pip (with spread) if you were to take a random entry going in any direction without care of patterns or direction or trend. I’m certain that with following my rules outlined, that the probability of success for a trade within 48hrs is greater than 99.9% and most likely approaching 99.99%. Don’t be greedy, go for 2-4 perfectly timed trades each day that are in line with the rules and you will have great success. Green pips to all
Thank you very much my friend. You have opened my eyes …. a bright light will soon come true ….
Hello,
Any advice on how to get around the problem of not being able to click Buy or Sell, due to the horizontal lines getting in the way.
On a 1 minute chart, I find it critical to get filled when I click Buy or Sell but so often find that I have clicked a Dotted Horizontal line.
Have you found a way around this ? Or is it possible to have an update where we can turn the Horizontal lines on/off ?
Or do I need to set my charts up differently ?
3 pip targets
Have you found a way around this ? Or is it possible to have an update where we can turn the Horizontal lines on/off ? Or do I need to set my charts up differently ?
yes, i will post a new version later with a few updates
Focus, Patience, Determination & Order in chaos
A question: Is it right that a broker can ‘do something’ to our EA, after they know that our EA brings us good profit ? How to prevent this? Thanks
I plan on getting the statistics for tick data which I downloaded recently. I post the charts when I get the chance. What it will look like is an exponential decay where as you increase the TP the probability of success drops exponentially 
- You must be logged in to reply to this topic.