Do Trendlines work?
- Thread starter duard
- Start date
What is the p>.05 referring to? A percentage of the monthly bar or a minimum movement?
At this point some rows can be added to enhance the five of six columns. The inputs to these rows will come from other places in the Forex realm.
Is this referring to other raw data? What other places are there?
By applying this beginning to the six most common pairs, the relationships of the pairs may be understood.
Is this a reference to using the template to understand trading this pair, or relationships/influences of each pair on another pair?
Fortunately the template is a familiar one.
Some questions on cyber-pace:The PV volatility matrix was based on volume to get an expectation for price. Volume was leading indicator of the expected volatility.
Using volatility, a distribution doesn't help unless it's based on something (tick volume for instance). Or it would be just divided into deciles and pace assigned that way. IOW, doing the reverse, using volatility to get an expectation of volume. Would then need a pro-rata volatility?
With bar overlap it's a similar situation (what basis to expect overlap without volume), though it could be used as a trending indicator similar to ADX or DMI measuring the bar extension or non overlap.
Or are we looking at combining both of these in a way to work out the pace?
Then use PRV (prorata volatility) for anticipation. This would get us to the "Bar Vol Expected" and "overlap expected" boxes.
Using the tick volume, volatility and adjusting with overlap could create a type of volume bar. Is this correct or going off on a tangent?
Tick volume seems like watching the T&S speed accelerate an decelerate. Which could be helpful as a fine control.
Coming to the CRT display, would this cyber-pace (if converted to a histogram) replace the tick volume on the display?
Would it be correct to assume the lower indicators are the standard template ones - Stoch, MACD.
Trend Reversion box: Are we reverting to the mean, or a trendline? I'll post questions on that box separately.
Sorry for all the questions, it's the hygiene worker, primitive community situation (Anopheles Mosquito - Shaw)
Regards - EZ
Quote from epetrov:
Ezzy,
you must be the only one tried to follow that boxes searching for some logic.
I thought I was the only who didn't try to follow
His chart was better than the boxes. The chart was easy to understand and the trend lines were well drawn. The problem I have with moving ave, and macd is that that most of these indicators tell you what already happened. However, some indicators may be useful to determine overall longer term trend.
Quote from tradersaavy:
I thought I was the only who didn't try to follow
In the vast majority of cases indicators provide confirmation and should not be used as a trigger. That said there are some (like cd23) who use such things as MACD and stochastics to anticipate more subtle price events.
Quite right that he knows how to draw a trendline which most of the participants in this thread don't. He uses the Sperandeo variant which, IMO, is the only one that when used correctly gives consistent results. Of course one must know what one is looking for and, IMO, without a left trendline, one is not making best use of the RTL. To ignore volume when assessing trendlineline (RTL and LTL) price action is puerile at best. A Sperandeo "1, 2, 3" is about as good as one can do with just an RTL.
lj
Ezzy partial quote: With bar overlap it's a similar situation (what basis to expect overlap without volume), though it could be used as a trending indicator similar to ADX or DMI measuring the bar extension or non overlap.
There is precious little of consequence in ET on the topic of bar overlap although clearly it is an important one. cd23, in this and past incarnations, has talked about this phenomenon and as well the more extreme situations of HVS (High Volatility Stalls)and CCC (Congestion, Convergence and Centering). So far, I've only been able to find this more general tabulation of pace and overlap correlation (in this case EOD):
http://www.elitetrader.com/vb/showt...age=6&highlight=jersey and ball&pagenumber=13
04-10-07 04:13 PM - second post from the top.
Here is the overlap chart for 2006 (see my attachment). Sometimes the days are different and some times they continue to follow the prior day. Here you can see how this works out by this comparison of adjacent bars. This will help lay more ground work (context) for scalping using the fine cluster of display items.
and this comment from PointOne later on in the same thread:
04-10-07 07:14 PM - second post from the bottom.
Just so I know if I'm reading the chart correctly, picking a point in the middle of your table at random: On 6 days in 2006 the overlap of EOD bars (relative to day before) was 7 points when the pace was defined as medium (forced curve). Negative overlap = gaps. The table shows that pace and bar overlap (a proxy for volatility) are proportional to some significant degree (if you hilight the days for each pace so you get more than 25 in each row using minimum columns you can see this).
It is always interesting how different people can look at a data set and see different and sometimes opposing things. To my eye there appears to be a drift to less overlap at lower pace but given that the volatility (= bar H - bar L) is anticipated to be less at lower pace, I don't know how useful this is. I don't see why overlap can be equated with volatility as P1 asserts. IMO we are talking about different entities. Overlap has to be looked at in terms not only of its extent but also with respect to the value of each bar's volatility. Perhaps a "normalizing" operation such as looking at the % or fraction of overlap might be of utility.
My observations are so simplistic that there must be more to this than what I am presently seeing. Your thoughts would be of interest.
lj
There is precious little of consequence in ET on the topic of bar overlap although clearly it is an important one. cd23, in this and past incarnations, has talked about this phenomenon and as well the more extreme situations of HVS (High Volatility Stalls)and CCC (Congestion, Convergence and Centering). So far, I've only been able to find this more general tabulation of pace and overlap correlation (in this case EOD):
http://www.elitetrader.com/vb/showt...age=6&highlight=jersey and ball&pagenumber=13
04-10-07 04:13 PM - second post from the top.
Here is the overlap chart for 2006 (see my attachment). Sometimes the days are different and some times they continue to follow the prior day. Here you can see how this works out by this comparison of adjacent bars. This will help lay more ground work (context) for scalping using the fine cluster of display items.
and this comment from PointOne later on in the same thread:
04-10-07 07:14 PM - second post from the bottom.
Just so I know if I'm reading the chart correctly, picking a point in the middle of your table at random: On 6 days in 2006 the overlap of EOD bars (relative to day before) was 7 points when the pace was defined as medium (forced curve). Negative overlap = gaps. The table shows that pace and bar overlap (a proxy for volatility) are proportional to some significant degree (if you hilight the days for each pace so you get more than 25 in each row using minimum columns you can see this).
It is always interesting how different people can look at a data set and see different and sometimes opposing things. To my eye there appears to be a drift to less overlap at lower pace but given that the volatility (= bar H - bar L) is anticipated to be less at lower pace, I don't know how useful this is. I don't see why overlap can be equated with volatility as P1 asserts. IMO we are talking about different entities. Overlap has to be looked at in terms not only of its extent but also with respect to the value of each bar's volatility. Perhaps a "normalizing" operation such as looking at the % or fraction of overlap might be of utility.
My observations are so simplistic that there must be more to this than what I am presently seeing. Your thoughts would be of interest.
lj