range vs trend and how to predict it

[jficquette]

Ok got to any of the yahoo link and click on the report and it will give you who important it is on a - d scale.

http://biz.yahoo.com/c/e.html

For instance, on construction spending:

Importance (A-F): This release merits a D.
Source: The Census Bureau of the Department of Commerce.
Release Time: 10:00 ET on the first business day of the month (data for two months prior).

So just go thru and learn the A's.

John

 

[apex82]

Sorry Razor, I was looking at the wrong chart. I use it on a lower timeframe as a CF and got them mixed up. Your chart data is still a bit different then mine though. Make sure you are using an EMA and the daily data, not the 24 hour. I'll post a chart for reference.

 

[flip]

Quote from Arnie:

I think you really need to look at intraday day data to find a stock or SIF really trended (i.e opended near H/L and closed near L/H). I am currently working on trying to measure the intraday volatility along these same lines of thinking. Just looking at the EOD data isn't enough. A stock could have a nice trend in the morning, then reverse with a nice trend in the other direction and close near the open (all in all a very tradeable setup) vs a stock that just gets whipped round all day, but the EOD data would look identical to the former.

that was my thought too, so i lately started to define trendiness using intraday data. i did some initial work with the “efficiency ratio”, also called “fractal efficiency”. perry kaufman uses this concept in his adaptive moving average. sounds complicated, but actually it’s rather straightforward. the sum of the absolute value of all single intraday movements from bar to bar (over a given time period, say 5min) is set in relation to the total movement of this day from open to close. if a day opens on the low and closes on the high, and each x-min bar (x = 5 in my case) is higher than the previous bar, the value would be 1. of course this is never the case, but the more directional intraday movement (hence the more trendiness), the higher this value.

of course there’s still the problem, that a day might have been trending overall but open = close, so the value is 0 again.

what’s your way of measuring trendiness using intraday data? did you come up with interesting results?

 

[travis]

John, thanks for the advice. I know you are right in saying that the trend is caused by the news. Yet, those people saying "it's all in the chart" are also right, because you said it yourself that price is influenced by the news. So if you look at price you see the news in it.

Now the objection of course will be that you have to wait until it's too late in order to see the news in the chart, whereas by knowing which news causes trend, you could forecast it. This is true. Knowing what news is a market mover will give you an edge, provided that you do know which news is and which is not.

Yet there is also a big disadvantage in watching the news. First of all, you could be wrong in predicting which news will be a market mover. Second of all, it's practically impossible to create an automated system (entirely automated) that gives you an estimate of future trend or range based on what news will be released, at what hour, and on which day. Or at least, personally, I can't create a system that browses the web, gathers information about dates and times of what news, and decides whether it's important.

In my opinion, for an automated system, observing pre-open or "first half hour" data is the only possible way to predict whether today will be trend or range. And it might be even more accurate, because of what I mentioned about the possibility of being wrong about judging the importance of news (whereas if nothing happens on the chart after the news is released you know for sure that the news was irrelevant).

Say that at 8.30 AM, as it often happens, important news is released and the market moves a lot. My system will pick it up by observing an unusually high pre-open range, and since it doesn't start trading until 9.30 anyway, it will be able to choose the right type of trading by that time. There might be a problem for the news released at 10.00. Maybe I will adapt it, and postpone trading until then. But probably for those few days that it happens, it's not worth changing the whole system.

 

[Razor]

Thanks John and apex for your repsonses.

Apex, I can't seem to get the ATR/EMA to look like yours. I don't think it will make that much of a difference though.

Thanks again

 

[jficquette]

Quote from travis:

John, thanks for the advice. I know you are right in saying that the trend is caused by the news. Yet, those people saying "it's all in the chart" are also right, because you said it yourself that price is influenced by the news. So if you look at price you see the news in it.

Now the objection of course will be that you have to wait until it's too late in order to see the news in the chart, whereas by knowing which news causes trend, you could forecast it. This is true. Knowing what news is a market mover will give you an edge, provided that you do know which news is and which is not.

Yet there is also a big disadvantage in watching the news. First of all, you could be wrong in predicting which news will be a market mover. Second of all, it's practically impossible to create an automated system (entirely automated) that gives you an estimate of future trend or range based on what news will be released, at what hour, and on which day. Or at least, personally, I can't create a system that browses the web, gathers information about dates and times of what news, and decides whether it's important.

In my opinion, for an automated system, observing pre-open or "first half hour" data is the only possible way to predict whether today will be trend or range. And it might be even more accurate, because of what I mentioned about the possibility of being wrong about judging the importance of news (whereas if nothing happens on the chart after the news is released you know for sure that the news was irrelevant).

Say that at 8.30 AM, as it often happens, important news is released and the market moves a lot. My system will pick it up by observing an unusually high pre-open range, and since it doesn't start trading until 9.30 anyway, it will be able to choose the right type of trading by that time. There might be a problem for the news released at 10.00. Maybe I will adapt it, and postpone trading until then. But probably for those few days that it happens, it's not worth changing the whole system.

I agree that using the reports that are released periodically throughout the month are pretty much impossible to automate. Plus my comments were related to daytrading and not using reports to swing trade etc.


I was just sharing my observation that when the market really trends there is something making it and that is usually some report or news or it has to do with the day of the month.

No magic bullet. I just know that if all it was to it was the charts then it would be easy and we all know that it's not.

I think most future traders try to trade everyday day. Everyday does not turn out to be a good day because a lot of days the market goes no where.

If one could identify those days in advance where there is more likely to be movement then one would trade less and be better positioned to take advangtage of trendy days if they happen.


John