Dick Lexic
Veteren member
- Messages
- 4,954
- Likes
- 124
i've got a conflict of signals ...the downside is 5900 cash..with 940's permission of course
jimi - fib numbers really only work because everyone is looking for them now. There might have been a time when they happened naturally, but now everyone is looking at the 38, 50, 62, 78, 127 and 162% levels. Especially on the 1 hour chart.
Split - Where did you close your long from the first hour? I closed at 48 at the top of the channel on the 5 min chart. I didn't expect a blow off top. I did go short at 58 but only hung on for a few points. Difficult to tell what to do when that happens, but I will make a note of what to do on a blow off top for next time.
Dick Lexic
Veteren member
- Messages
- 4,954
- Likes
- 124
tight squeeze at the mo...
Someone asked me, I think Martin or Split, if my '(simple) automated rule-based system' worked better on certain days of the week. The answer is No.
I looked at the 'day of the week' analysis of profits and losses by quarter for 2011 & 2012. (The system makes several trades per day).
It would have been nice if one day of the week always generated losses because the remedy would be: don't work Wednesdays, or Monday and Fridays, or whatever.
The analysis is NOT statistically significant and there are dozens of other things to consider, but I'll not be following it up. If certain days of the week exhibited certain behaviours, then it would be common knowledge. But that's not the point at all, is it? It's whether my system with specific parameter values has day of the week bias. Of course, it would be possible to optimise the parameters to give the result I wanted, but that's the problem with optimisation. What is needed is a 'robust set of parameters' that are not 'specific event' dependant. Some of you guys have been asking about the Turtle book, well it goes into similar ideas in some detail.
Like you, I’m waiting for something to happen today, so I’ll just chat for a while. Read it if you want, or not. I can give you my guesses for what might happen, but that's all they are. The past, however, I can describe in great detail. Yesterday the high of 5990 was tested 3 times a 9:20, 11:00 and 2pm. Then the price dropped below the previous low of 5970 at about 3pm. (Times and prices are approximate). When prices test a level 3 times and give up, there’s likely to be a strong movement in the opposite direction. Isn’t there? Yesterday and today a low of 5932 has been tested and today we have what looks like a head and shoulders. There could still be a decent move today. I think it’s going to be Up … Watch it go down!
I looked at the 'day of the week' analysis of profits and losses by quarter for 2011 & 2012. (The system makes several trades per day).
It would have been nice if one day of the week always generated losses because the remedy would be: don't work Wednesdays, or Monday and Fridays, or whatever.
The analysis is NOT statistically significant and there are dozens of other things to consider, but I'll not be following it up. If certain days of the week exhibited certain behaviours, then it would be common knowledge. But that's not the point at all, is it? It's whether my system with specific parameter values has day of the week bias. Of course, it would be possible to optimise the parameters to give the result I wanted, but that's the problem with optimisation. What is needed is a 'robust set of parameters' that are not 'specific event' dependant. Some of you guys have been asking about the Turtle book, well it goes into similar ideas in some detail.
Like you, I’m waiting for something to happen today, so I’ll just chat for a while. Read it if you want, or not. I can give you my guesses for what might happen, but that's all they are. The past, however, I can describe in great detail. Yesterday the high of 5990 was tested 3 times a 9:20, 11:00 and 2pm. Then the price dropped below the previous low of 5970 at about 3pm. (Times and prices are approximate). When prices test a level 3 times and give up, there’s likely to be a strong movement in the opposite direction. Isn’t there? Yesterday and today a low of 5932 has been tested and today we have what looks like a head and shoulders. There could still be a decent move today. I think it’s going to be Up … Watch it go down!
jimi - thanks for the follow up on your system for days of the week. There must be a way to optimise it depending on if its ranging or not. Maybe something to do with daily range size compared to the ATR or use ADX when ADX is higher than AD+ and AD-. If you watch ADX you will know this. When the ADX is below AD+ and AD- then it is range bound, but when it makes it above them then it is trending. You might have to experiment with time frames, but maybe the hourly or 15 min or 5 min depending on your time horizon.
As for today, it all boils down to the options market and tomorrows expiry. Also watch if AAPL hits 600. Sure to bounce there.
As for today, it all boils down to the options market and tomorrows expiry. Also watch if AAPL hits 600. Sure to bounce there.
I'm sorry, Martin, I'm on cash. I must remember to say that. It must have been about 0940 because I wrote a post and went shopping without looking anymore.
Dick Lexic
Veteren member
- Messages
- 4,954
- Likes
- 124
jimi, do you include volume in your system ?
maybe rather than a specific day you should try a cycle ...like the Taylor cycle. if you identify the buy day for instance then as taylor said don't change it...maybe something might show up in your back test...just a thought
barjon
Legendary member
- Messages
- 10,752
- Likes
- 1,863
Just for interest's sake FTSE has its weakest day of the year yesterday in comparison to DOW at -65 by the close (4:30) - ie: should have closed up at 6010 on the cash had it kept pace with dow.
It's continued in the same vein today and is currently a further -28 weak. Seems it's itching to go down and only holding where it is by the DOW - and then very reluctantly.
needless to say, I'm long ftse/short dow from early on and well underwater atm
It's continued in the same vein today and is currently a further -28 weak. Seems it's itching to go down and only holding where it is by the DOW - and then very reluctantly.
needless to say, I'm long ftse/short dow from early on and well underwater atm
Dick Lexic
Veteren member
- Messages
- 4,954
- Likes
- 124
How to go insane in one easy lesson.
Daytrade tade the stock market
Daytrade tade the stock market
No, it's FXCM's 'Strategy Trader'. It seems that their intention is to stop supporting the platform and replace it with MT4. Sad.
I guess not good when you have something up and running. But MT4 should do the same trick.
Just for interest's sake FTSE has its weakest day of the year yesterday in comparison to DOW at -65 by the close (4:30) - ie: should have closed up at 6010 on the cash had it kept pace with dow.
It's continued in the same vein today and is currently a further -28 weak. Seems it's itching to go down and only holding where it is by the DOW - and then very reluctantly.
needless to say, I'm long ftse/short dow from early on and well underwater atm
jon - the ftse has been incredibly weak these last two days. I dont know which componet is causing it, or if it's just an options trader protecting his position. It should all get back to normal on monday.
Just to add to that. It was weak leading up to the FOMC statement. Then after that the DOW has rocketed, with a brief drop after the failed stress test news which was very short lived.
Jon, we have been weak all year, but as this was an extreme move, have you got any data to support some sort of correction?
barjon
Legendary member
- Messages
- 10,752
- Likes
- 1,863
well it might feel like it, reksa, but at the close yesterday it was only -74 weak from where it started the year - and we had -65 of that yesterday!! it's been up and down - at best it got to +64 strong but that was only a couple of days into the year. It gradually deteriorated from there until it got to about -30 weak, then picked up to get to around +50 strong mid-February, then a gradual decline which picked up pace in the last week or so to the -74 to yesterday.
jon
