Predicting The News | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Predicting target price?????? 4 replies

Predicting the news 0 replies

Only now I see a lot of Traders predicting the USD rallie for 2008! 2 replies

Predicting direction of economic releases 123 replies

Pure Price Action: Predicting the Current Move 9 replies

Attachments: Predicting The News
Exit Attachments

Predicting The News

  • #1
  • First Post: Feb 29, 2008 3:10am Feb 29, 2008 3:10am
Hello TR folk... I have a question to all of you out there:

Have any of you played around with predicting the news? That is, making your own estimates for economic indicators?

I saw the NFP Sweepstakes thread in one of my searches, but I don't think that really counts

I'm wondering especially if anyone has tried using a combination of published indicators released earlier within a month to predict the values of those later that month. It seems that this might be a problem quite appropriate for neural networks (your biological ones or my artificial ones) or other statistical methods.

Does anybody know an easy way to gather the past values for nearly all economic indicators?
  • #2
  • Feb 29, 2008 7:55am Feb 29, 2008 7:55am
It can be done. After all, every event is linked to some previous event. Lagging economic indicators (fundies) can be "predicted."

The question is, can it be used?

e.g A good NFP report can push the greenie lower or a bad CPI out of Japan can push the yen against all majors. Predicting the numbers for me is rather less significant than predicting the market re-action to the numbers.

Economist, they work on positive statements. As a trader the person must work with normative stuff where his decisions has to be made based on subjective judgment of the value equations.

Just my humble opinion.
  • #3
  • Edited 9:12am Feb 29, 2008 8:08am | Edited 9:12am
Quoting drummer486
Hello TR folk... I have a question to all of you out there:

Have any of you played around with predicting the news? That is, making your own estimates for economic indicators?

I saw the NFP Sweepstakes thread in one of my searches, but I don't think that really counts

I'm wondering especially if anyone has tried using a combination of published indicators released earlier within a month to predict the values of those later that month. It seems that this might be a problem quite appropriate for neural networks (your biological ones or my artificial ones) or other statistical methods.

Does anybody know an easy way to gather the past values for nearly all economic indicators?
The problem with answering this question is that the method by which one forcasts a news event is by it's own definition an inefficiency. If word got out how it was done then the loop hole would be plugged up. Having said that, I have forcasted correctly the direction, appx amount and magnatude of every NFP since April of last year and have profited by placing trades at least 12 hours in advance of it. Merlin can confirm this.

So yes, the data is approximately available in other sources. The thing is one does not need to know exactly, only if there is something unexpected to come out. Other wise the news event playing out at it's normal time will effect the markets or not, as it would have with out knowing the forcast.

Sadly, on a side note, the reason the NFP has lost much of it's punch is in my oppinion, because others have also discovered this inefficiency. This is how these things work. They only work until enough people know about them and then they become diluted.

Addendum: Just wanted to add, when I think about it, it is not a forecast or a prediction. It is more like a rooting out of the same facts from other sources. The info was already cast, just not released yet.
  • #4
  • Feb 29, 2008 11:12am Feb 29, 2008 11:12am
Yeah, I suppose this thread creates somewhat of a conflict of interest. It is encouraging to hear that it can successfully be done, however. smjones - did you do this by intuition or did you use some kind of calculation?

And again, do you know an easy way to download historical values for all releases? I might just build a little script to read all the old FF calendars, but I was hoping to avoid this.
  • #5
  • Feb 29, 2008 11:35am Feb 29, 2008 11:35am
Quoting drummer486
Yeah, I suppose this thread creates somewhat of a conflict of interest. It is encouraging to hear that it can successfully be done, however. smjones - did you do this by intuition or did you use some kind of calculation?

And again, do you know an easy way to download historical values for all releases? I might just build a little script to read all the old FF calendars, but I was hoping to avoid this.
It is a claulation based on other historical data. I would like to be more specific, but it is still working for me, so I will keep using it until it does not, then I'll write an ebook and put it on ebay. No just kidding... LOL

I would imagine it's days are numbered though, because I have seen a smoothing out of the effect since I started doing it last year.

Which is another thing you may consider. Anything you are able to create from the data, can be created by others and will eventually bring about it's own demise. Such is the nature of these types of inefficiencies in markets.
It is just a matter of time.

In regards to FF calander, Yes, I used some of that data, but I am only average in programming, so I would not know how to use the data there in a database for your inputs.

BTW, call me Scott
  • #6
  • Feb 29, 2008 1:43pm Feb 29, 2008 1:43pm
predicting news releases is as old as dow theory...have you ever wondered where those "forecasts" on calendars come from? basically the large news organizations like bloomberg and reuters have these polls setup - they regularly poll a number of economists on what they expect the upcoming number to be. each of these economists has their own method for predicting the numbers, but most base it on data that comes out earlier. for instance, NFP is the first new data to be released each month...economists will study this NFP report to help them predict CPI and GDP. this is just a simple example, but it illustrates how the process works. anyway, these polls are done and the average is taken, and viola, we have a forecast.

now you can see why the timing of the data makes all the difference. the fresher the data, the more important, because this data is used to predict future releases.

so you might think this is a game played only be the smart economists of the world, but as scott pointed out, all of us can play! many months ago me and scott were discussing the correlation of ADP and NFP (ADP comes out 2 days earlier and is based on the same month). both of us came up with our own models for the correlation and both have been making money from it. unfortunately these inefficiencies only last a short period because traders catch on quick these days, and as scott mentioned this ADP/NFP inefficiency is already drying up a bit. she still has some juice left but not much!

great topic btw
Relax and be happy.
  • #7
  • Feb 29, 2008 9:01pm Feb 29, 2008 9:01pm
Quoting drummer486
Hello TR folk... I have a question to all of you out there:

Have any of you played around with predicting the news? That is, making your own estimates for economic indicators?

I saw the NFP Sweepstakes thread in one of my searches, but I don't think that really counts

I'm wondering especially if anyone has tried using a combination of published indicators released earlier within a month to predict the values of those later that month. It seems that this might be a problem quite appropriate for neural networks (your biological ones or my artificial ones) or other statistical methods.

Does anybody know an easy way to gather the past values for nearly all economic indicators?

Here's a couple of places to look for the data:

http://www.rogerseconomics.com/Indic...DataSites.html

http://economagic.com

In the archives on my site there's a research project dated February of 2007 where I studied the effects of releases on prices. That's when I found the above two sites. I actually ended up extracting data from Forex Factory's calendar however since I wanted the forecast series too and no one else had that.

I haven't done anything yet on predicting the releases themselves, but my approach would be some kind of economic model based on the actual real-world relationships.
  • #8
  • Mar 1, 2008 10:02am Mar 1, 2008 10:02am
Quoting merlin
predicting news releases is as old as dow theory...have you ever wondered where those "forecasts" on calendars come from? basically the large news organizations like bloomberg and reuters have these polls setup - they regularly poll a number of economists on what they expect the upcoming number to be. each of these economists has their own method for predicting the numbers, but most base it on data that comes out earlier. for instance, NFP is the first new data to be released each month...economists will study this NFP report to help them predict CPI and GDP. this is just a simple example, but it illustrates how the process works. anyway, these polls are done and the average is taken, and viola, we have a forecast.

now you can see why the timing of the data makes all the difference. the fresher the data, the more important, because this data is used to predict future releases.

so you might think this is a game played only be the smart economists of the world, but as scott pointed out, all of us can play! many months ago me and scott were discussing the correlation of ADP and NFP (ADP comes out 2 days earlier and is based on the same month). both of us came up with our own models for the correlation and both have been making money from it. unfortunately these inefficiencies only last a short period because traders catch on quick these days, and as scott mentioned this ADP/NFP inefficiency is already drying up a bit. she still has some juice left but not much!

great topic btw
i agree, there's no longevity in news trading, trust me i know, if someone was to develop a system for news trading i will tell them make all you can make from it while you can and don't make a career out of it because it won't last long
If I Go Broke Trying Then I Will die happy.
  • #9
  • Mar 1, 2008 11:52am Mar 1, 2008 11:52am
Quoting Moe
i agree, there's no longevity in news trading, trust me i know, if someone was to develop a system for news trading i will tell them make all you can make from it while you can and don't make a career out of it because it won't last long
Moe, I'm curious as to why you'd say this. I can understand that spike trading is a dying art, especially as computer technology improves and brokers learn how to better deal with the inefficiencies. I don't see why econometric forecasting would become obsolete, however. If one could develop a system that could accurately forecast news releases, one could open and close trades during normal market conditions, avoiding the fast markets of the spikes.

Or do you think that more people will be focusing on econometric predictions, so that any prediction system would have so much competition that it would lose its edge?
  • #10
  • Mar 1, 2008 11:58am Mar 1, 2008 11:58am
Firstly let me say that I don't trade the news and for the life of me can't understand why anyone else bothers to try.

But predicting the news is not the important factor.

What is - is predicting the reaction to the news.

Those are two completely different considerations.

I.E. Bad news sometimes is good and good news is sometimes bad.

I can figure it out some of the time and some of the time not, which to me equals a coin toss. The "experts" wish they were 50% accurate. But I strive for better than 50/50 before I would put hard earned money on the line.
  • #11
  • Mar 1, 2008 2:15pm Mar 1, 2008 2:15pm
Quoting drummer486
Moe, I'm curious as to why you'd say this. I can understand that spike trading is a dying art, especially as computer technology improves and brokers learn how to better deal with the inefficiencies. I don't see why econometric forecasting would become obsolete, however. If one could develop a system that could accurately forecast news releases, one could open and close trades during normal market conditions, avoiding the fast markets of the spikes.

Or do you think that more people will be focusing on econometric predictions, so that any prediction system would have so much competition that it would lose its edge?
one of the reason i say that is because back in the day the news react according to the news/numbers and so on, now how every the news comes out the market goes it's own way, i think thats because to many traders was predicting the news, you can predict the news but the market won't react according to the news, like i say don't make a
career out of it, in the end you will get burn.
If I Go Broke Trying Then I Will die happy.
  • #12
  • Mar 1, 2008 6:36pm Mar 1, 2008 6:36pm
Quoting Moe
one of the reason i say that is because back in the day the news react according to the news/numbers and so on, now how every the news comes out the market goes it's own way, i think thats because to many traders was predicting the news, you can predict the news but the market won't react according to the news, like i say don't make a
career out of it, in the end you will get burn.

I will vouch for that one.......like Suntrader said the key is the "REACTION" to the news.

HOLLA
  • #13
  • Mar 1, 2008 7:17pm Mar 1, 2008 7:17pm
Quoting SunTrader
But predicting the news is not the important factor.

What is - is predicting the reaction to the news.

Those are two completely different considerations.
Bingo
Patience + Humility + Study = Success
  • #14
  • Mar 1, 2008 8:44pm Mar 1, 2008 8:44pm
Quoting SunTrader
Firstly let me say that I don't trade the news and for the life of me can't understand why anyone else bothers to try.

But predicting the news is not the important factor.

What is - is predicting the reaction to the news.

Those are two completely different considerations.

I.E. Bad news sometimes is good and good news is sometimes bad.

I can figure it out some of the time and some of the time not, which to me equals a coin toss. The "experts" wish they were 50% accurate. But I strive for better than 50/50 before I would put hard earned money on the line.
im glad you bring this up ( predicting the reaction to news ) this is why news trading won't last long, like i said before the reaction is not a consiting thing it change from time to time the news might react according to the way you predict it lets say 3 times in a row then just when you think you got it, it start reacting diffrent again and this is something you can't predict.
If I Go Broke Trying Then I Will die happy.
  • #15
  • Mar 2, 2008 3:34pm Mar 2, 2008 3:34pm
Quoting Moe
there's no longevity in news trading
ive been "news trading" for most of my career, it is a very profitable area to focus on.

i believe you are referring to "spike trading", which is the newbie way of trading the news. unfortunately, when people say "news trading", most traders think of spike trading. i have never spike traded, and i agree there is no longevity in it due to the lack of liquidity when the data hits the wires.

news trading to me is trading the run-up-expectations or reaction to news releases (pre-scheduled or otherwise). most of my news trading strategies are based on a very simple concept: in a bull market, fade bad news when the price action tells you its safe to do so (and vice versa).

btw moe, didnt your "mentor" spike trade successfully? i vaguely remember you mentioning something about that.
Relax and be happy.
  • #16
  • Mar 2, 2008 3:38pm Mar 2, 2008 3:38pm
Quoting Moe
now how every the news comes out the market goes it's own way
the market has always been like that, i cant remember any extended period where the market actually reacted the way it "should".
Relax and be happy.
  • #17
  • Mar 2, 2008 4:03pm Mar 2, 2008 4:03pm
Quoting merlin
ive been "news trading" for most of my career, it is a very profitable area to focus on.

i believe you are referring to "spike trading", which is the newbie way of trading the news. unfortunately, when people say "news trading", most traders think of spike trading. i have never spike traded, and i agree there is no longevity in it due to the lack of liquidity when the data hits the wires.

news trading to me is trading the run-up-expectations or reaction to news releases (pre-scheduled or otherwise). most of my news trading strategies are based on a very simple concept: in a bull market, fade bad news when the price action tells you its safe to do so (and vice versa).

btw moe, didnt your "mentor" spike trade successfully? i vaguely remember you mentioning something about that.
yes he did so did i but things did change i got out alive and made good money doing it, and btw you have always traded the fade, am i correct?
If I Go Broke Trying Then I Will die happy.
  • #18
  • Mar 2, 2008 4:15pm Mar 2, 2008 4:15pm
well glad to know you got out alive, thats all we can ask of our learning experiences

and yeah, ive been known as a fader not because of some personality thing, i just think fade situations are easier to predict than momentum, in short term trading at least.
Relax and be happy.
  • #19
  • Edited 11:52pm Mar 2, 2008 7:11pm | Edited 11:52pm
Very interesting thread. Firstly because of this "predcting the numbers" thing, and I definitely see where merlin and scott are coming from there, although I have never attempted to spend time on this myself for the reason scott mentioned: inefficiencies get plugged pretty fast. Way to go for making the discovery though, we'll all stay hush about it

Second, the actual reactions to the news. I've always been intrigued about how merlin and others identified themselves as "news" traders while others and myself barely even look at the events calendar (despite significant hell-raising on cesars hangout thread ).

So merlin : can you clarify what exactly you mean by "fading"? Do you mean scalping pullbacks after the initial rage of the releases? Is there a thread about this anywhere?

Cheers,
-A
Virtue finds and chooses the mean. Aristotle, Ethica Nichomachea
  • #20
  • Mar 2, 2008 7:41pm Mar 2, 2008 7:41pm
Quoting merlin
the market has always been like that, i cant remember any extended period where the market actually reacted the way it "should".
Absolutely.
0 traders viewing now