Multi-indics indicator - page 24
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
Have taken 29 pips on EURUSD today
I need to backtest it to have full confidence in it, but I believe that this has a chance to be a high probability system. Currently I only put 1 microlot trades to get the feel for it.
Ferru, Have a look at my thread https://www.mql5.com/en/forum/177151
where I investigate the relations between currencies and their strenghts. I've made a few excel sheets over a few years of daily data. I too thought that trading in the direction of the strongest currency against the weakest one would bring profit. It is logical, but in fact, as you look at the ixmeter readings, the currencies achieve the levels after a substantial move, and therefore you have very little move left which cuts your probability of success. Moreover, I found with my sheets that the currencies don't have a tendency to follow through with the strenght immediatelly the next day, and the results are pretty much 50/50 and that lends it unpredictable. I couldn't find any more use with the stats I made so I dropped it for now until I come up with another idea or someone very smart tells me how I can use it.
Regarding your multi-indi, I think you are onto something very useful, and I will have to find the time to backtest the stochastic entry, because so far I see that the combination of two will keep you out of a ranging market. The reason is that when your indicator moves along with the price, and the price pulls it into the other direction (from buy to sell and vice versa), the stochastics shows the opposite entry. only if the trend is in place, the retrace in the trend will pull the stochastics far enough for us to enter but it won't change your indicator's reading and we'll have an entry. Hope that makes sense.
Very glad to have met the two of you.Thanks Daniel for your interest. I look asap at your excel sheets and study what we can do with.
FerruFx
There was another entry on usdjpy at 12:45 GMT. Spot on for 30 pips. The markets are in a nice trend today.
edit: I've started looking through past events. For instance, cable on the 1st of June produced at least 7 stochastics signals, only two of them were backed by Trend, and one of them got stopped out - break even on cable that day. This is only to tease you guys
more testing to come.
edit: another thing I'd consider berofe entering is the extent the pair has moved already in the day - how much of the average daily move it has covered.
Ferru, Have a look at my thread https://www.mql5.com/en/forum/177151
where I investigate the relations between currencies and their strenghts. I've made a few excel sheets over a few years of daily data. I too thought that trading in the direction of the strongest currency against the weakest one would bring profit. It is logical, but in fact, as you look at the ixmeter readings, the currencies achieve the levels after a substantial move, and therefore you have very little move left which cuts your probability of success. Moreover, I found with my sheets that the currencies don't have a tendency to follow through with the strenght immediatelly the next day, and the results are pretty much 50/50 and that lends it unpredictable. I couldn't find any more use with the stats I made so I dropped it for now until I come up with another idea or someone very smart tells me how I can use it.
Regarding your multi-indi, I think you are onto something very useful, and I will have to find the time to backtest the stochastic entry, because so far I see that the combination of two will keep you out of a ranging market. The reason is that when your indicator moves along with the price, and the price pulls it into the other direction (from buy to sell and vice versa), the stochastics shows the opposite entry. only if the trend is in place, the retrace in the trend will pull the stochastics far enough for us to enter but it won't change your indicator's reading and we'll have an entry. Hope that makes sense.
Very glad to have met the two of you.Hi Daniel,
You have hit the 'nail on the head' in respect of the lack of benefit that the !xMeter is to the day trader, as an indicator of immediate direction for entry. Sure, it correctly shows you the stregnth that a particular currency is achieving over it's partnered pair, but by the time it is in it's high range of the scale at value 6-8 to be meaningful, the best part of the move is over and you are entering too late. It certainly may be useful in alerting you to a possible trend forming within a particular pair and the direction it is taking, that can be a very useful 'heads-up' for perhaps a trade after the first pullback.
For that reason, I cannot get enthusiastic over the !xmeter, as a primary signal indicator (as it is extremely lagging) and why I pointed out in my earlier post to FerruFx that I believe the MTF indy is far superior for this task. This is my personal opinion only. I know you must have put a tremendous amount of time, ideas and effort into your own thread in your pusuit of this approach and perhaps you are onto something, who knows?
The day trader who is only looking for 20-30 pips daily, but every day consistently, needs to look at and respect what the M5 - H1 charts are telling you, backed by the H4 for direction-of -the-moment (never go against the H4). The H1 and H4 stochastic and MACD (and a couple of MA's if you wish) are all that are needed to tell you which direction price is likely to take over the next few hours. The D1 is of little or no use to the day trader.
Your idea to backtest the indy, is of course very sound and I wish it was possible, but unfortunately, as it stands that cannot be done (see FerruFx's recent post). It is a pity, as it does need a backtest facility to achieve the best results.
johnny
It has a test facility, but only on a one-candle-at-a-time basis. It's time consuming, so it will be a while before I can post any conclusions. In fact I haven't started a proper test yet. I'm thinking on its form.
It obviously would have been easier if the output was in a graph form. Perhaps Ferru will be able to make such a version. Can you? A scale of 0 - 100% with a middle of 50%
edit: I've been going through 5 min cable history. So far I've made several observations:
Trade between 6am - 4 pm GMT (old truth)
daily profit 20-30 pips
Don't trade if daily average has been reached or is getting close
Let the trade go until the second attempt of stochastics going to overbought/sold area to reach around 20-30 pips (simple three wave trade)
If our trade starts losing and there is a signal in the opposite direction - reverse. it has a higher probability to succeed.
Interesting thing happened on 23rd may. Cable did over 150 pips but there was no entry into it. Trend was too strong, and before it started there was a fake move down - but without an entry as well. I switched to eurusd out of curiosity, and presto - around 30 pips net that day
More testing to come. Will keep testing forward with 1micro each trade as time permits.
edit 06.06.2007 After a few days of trend the pound and euro are up for a retrace. As it was to be predicted the readings of trend indicator were obviously not acurate enough during the reversal and it gave long signals for trades that eventially got stopped out. Two signals on cable one on euro, a short on swissy is still on. no entry for usdjpy. Basically a typical reversal/retrace day with surprises
edit: closed usdchf at 0 pips - going nowhere and stoch has been in overbought twice.
edit: usdjpy gave a short entry. closed at 17 pips after stoch went into oversold
See all the updates above. no more trading today I think, and I'll have to go to work too.
Have anybody notest that it don't work right in backtest?
Only the timeframe selected in backtest changes.
I agree. I was watching the four majors all day yesterday and saw the signals as they came. When later in the day I returned them, there were signals along with stochastics that weren't present in real time.... If there is a way to present the indicator in the form of graph, this would muchly simplify things.
If there is a way to present the indicator in the form of graph, this would muchly simplify things.
I am not sure but i look at this when i have more time! But the idea is interesting
FerruFx
If there is a way to present the indicator in the form of graph, this would muchly simplify things.
Something like this ...
For say the truth, i work on it since some time but it wasn't a priority for me.
(see attached)