Predicting The News | Page 3 | 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

  • #41
  • Mar 4, 2008 5:27pm Mar 4, 2008 5:27pm
Quoting drummer486
If one could accurately predict what the reaction to a given set of news releases would be, one could place a set of limit sells at the top 10% of the projected spike (or limit buys at the bottom 10% as appropriate). Then just wait for the spike to take out those orders and subsequently retrace. Has anyone attempted to set up an automated fading trade-placer based upon actual numbers as they come out?
i call this method a "reverse straddle" and i like it because it uses limit orders. i will tell you about a very similar idea that i learned right here on the FF forums... when the really big "whippy" releases come out, like interest rate statements, try using the reverse straddle on pairs that do NOT include the currency releasing the report. for instance, if a USD release is coming out, put a reverse straddle on the GBPJPY. ive had limited success with this method, obviously the problem i ran into is getting filled on my limits. i never gave this strategy my full devotion due to other priorities, but i do believe with some study it could be profitable.
Relax and be happy.
  • #42
  • Mar 4, 2008 5:31pm Mar 4, 2008 5:31pm
Quoting smjones
I was looking at mike's chart. you can see after the fade short is where the real money seems to kick in. I really don't think they played in this at all, other than to get a better price. just my 2cents.
yeah i agree. the other force at work, especially for that release, is that the small traders who got long at release time chickened out when the price stopped moving in their direction, thats why we saw that sharp spike down when price hit their stop losses (thats just my theory, of course).
Relax and be happy.
  • #43
  • Mar 4, 2008 5:37pm Mar 4, 2008 5:37pm
Quoting merlin
yeah i agree. the other force at work, especially for that release, is that the small traders who got long at release time chickened out when the price stopped moving in their direction, thats why we saw that sharp spike down when price hit their stop losses (thats just my theory, of course).
yes exactly and the big guys aren't just willy nilly clicking buttons. They need someone to provide the liquidity to their huge orders. That is of course what the retail traders job is, and these stop losses = liquidity too. At least this is what I have been taught, even with a trillion dollar market, there is never enough liquidity out there. and for those aware of some of what DS does, you can see it happen all the time especially during news
  • #44
  • Mar 4, 2008 5:40pm Mar 4, 2008 5:40pm
Quoting mbqb11
yes exactly and the big guys aren't just willy nilly clicking buttons. They need someone to provide the liquidity to their huge orders. That is of course what the retail traders job is, and these stop losses = liquidity too. At least this is what I have been taught, even with a trillion dollar market, there is never enough liquidity out there. and for those aware of some of what DS does, you can see it happen all the time especially during news
And even more so just before the news when the liquidity really disappears.
That's when the market can be pushed around.
  • #45
  • Mar 4, 2008 5:42pm Mar 4, 2008 5:42pm
Quoting smjones
And even more so just before the news when the liquidity really disappears.
That's when the market can be pushed around.
for sure!, but I do pull my orders on really big releases that have extremely uncertain outcomes(like the recent rate cuts), things get way to crazy then

During the normal day in and day out news(even big news), there is much more "order" to the chaos . From my experience thus far.
  • #46
  • Mar 5, 2008 2:49am Mar 5, 2008 2:49am
Quoting merlin
i call this method a "reverse straddle" and i like it because it uses limit orders.
That's exactly what I was going to call it!

Quote
obviously the problem i ran into is getting filled on my limits
Trouble getting filled on limits?? You should always get filled on limits as long as the price goes past them. That comes directly from the definition of a limit order. If a broker/ECN is filling others' orders at prices above that of your limit order, they should be tarred and feathered... or at least the IT guys who developed their trade-matching system should.
  • #47
  • Mar 5, 2008 2:50am Mar 5, 2008 2:50am
I purchased a temporary subscription to the news analysis program I mentioned earlier. Time to put my money (and time) where my mouth is. I suppose I should start with the already-proven ADP-NFP correlation. I'm sure predictions would be of interest to traders this week.

I'd like to approach analysis from a couple different angles. The first is simple regression analysis. With modern statistical analysis programs, it's fairly easy to determine which news releases are most predictive of a certain other release. It is possible that the correlations are non-linear, however. In the case that such a correlation appears, perhaps it might be necessary to use NNs for analysis among the releases that are most predictive of a certain other release.

I'll try to test these methods of analysis on the task of NFP prediction. I'll post updates in this thread as I make progress, and try to detail the steps involved in this process for you. In the meantime, keep up the discussion. I'm really enjoying reading it
  • #48
  • Mar 5, 2008 11:11am Mar 5, 2008 11:11am
Quoting drummer486
That comes directly from the definition of a limit order. If a broker/ECN is filling others' orders at prices above that of your limit order, they should be tarred and feathered...
come on man, you are talking to merlin-limit-order-jeffries here!

what i meant was that the spikes are usually so temporary that my limit orders got ignored (ie not filled). and btw, its not true that your limit must be filled if the price goes past your order level...brokers follow a FIFO system so if the liquidity gets dried up in an instant there is nothing left to fill me with. this is pretty normal.

Quoting drummer486
I'll try to test these methods of analysis on the task of NFP prediction. I'll post updates in this thread as I make progress, and try to detail the steps involved in this process for you.
looking forward to it!
Relax and be happy.
  • #49
  • Mar 5, 2008 3:03pm Mar 5, 2008 3:03pm
Quoting merlin
come on man, you are talking to merlin-limit-order-jeffries here!

what i meant was that the spikes are usually so temporary that my limit orders got ignored (ie not filled). and btw, its not true that your limit must be filled if the price goes past your order level...brokers follow a FIFO system so if the liquidity gets dried up in an instant there is nothing left to fill me with. this is pretty normal.
I'm not sure I understand this point... In my understanding, the Ask price is a representation of the lowest available limit order price. For the price to move above your sell limit order, open selling interest at that level must be exhausted. That is, your order must be either filled by the broker or cancelled by you. Otherwise, the platform is giving buyers an ask price that is not actually the lowest available!

I suppose this reinforces my preference for ECNs over DD brokers. It seems totally non-sensical to me that anyone's buy order should be filled at a price higher than the lowest available selling price. I guess that DDs don't necessarily have to match customer orders against each other, and thus can simply feed customers from their bank's liquidity, and fill as they see fit. I'm just so used to the open-market system of ECNs, where my limit sell order at 1.8999 should ALWAYS be filled before anyone else's limit sell order at 1.9000.

Or am I missing something?
  • #50
  • Mar 5, 2008 3:16pm Mar 5, 2008 3:16pm
Quoting merlin
yeah this is pretty much my requirement to trade anytime! i refuse to gamble with illiquid markets. i stick strictly to limit orders with reasonable spreads - otherwise i find myself sad and complaining about how unfair brokers are LOL.
Good one
Patience + Humility + Study = Success
  • #51
  • Mar 5, 2008 5:15pm Mar 5, 2008 5:15pm
Quoting drummer486
Or am I missing something?
no no i got confused and mispoke there, sorry thanks for catching that!

yes the price can come to my limit order but cannot go past it. for instance if i am trying to short 10 lots, i might not get filled on any of them, while other traders who had their order in before me would get filled (all at the same price of course). another factor here is that market orders take precidence over limit orders, no matter who has their order in first. by the time the price reaches my order, liquidity is gone and i am lucky to get a partial fill. the "game" to this strategy is to put your straddle close to the current price, but far enough so that there is room for profit. if your order is too far you end up with a non-trade, which is better than a loss so i tended to lean that way when trading it.
Relax and be happy.
  • #52
  • Mar 5, 2008 5:55pm Mar 5, 2008 5:55pm
i saw on another thread that one guy started to have members sharingthere own regions current economic news

for eg AUSTRALIAS GROWTH slowing SO the next planned interest rate in may will not be comming,
the currentinterest rate of 7.5 % just anouced yesterday

it is licklyto remain at this level.
the employment rate is going strong needing further oversees workers due to the strong minning sector,

the housing sector is lickly to slow due to the home afordability and high interest rates and home prices,

alreadythe banking sector is starting to feel the strain and are raising the interst rates above the resrve banks rate.

these facts were discussed on our morning news.

if we all shared these economic weekly news from our region it would help establishour knowledge of each currency economic news

any way thats just an idea i have for sharing easily our economic news

FROM Brian
  • #53
  • Mar 8, 2008 10:44pm Mar 8, 2008 10:44pm
I've decided to shift my focus away from news prediction (for now) and back to price analysis with NNs. I might include some econometric inputs to the price analysis NN to see if the they add predictive value (at least for longer timeframes).

In the meantime, if you're interested in the application of neural networks for econometric forecasting, you might check out the following paper:

http://www.smartquant.com/references...ks/neural6.pdf

For those of you that don't have time to read it, an interesting point they make is that forecasting the next couple months is very difficult, but the general long-term trend (6-12 months in the future) is much easier. Nothing we didn't really already know, but the rest of the paper has some interesting information for those who like the math part.
Attached Image
  • #54
  • Mar 9, 2008 1:12am Mar 9, 2008 1:12am
Quoting drummer486
I've decided to shift my focus away from news prediction (for now) and back to price analysis with NNs. I might include some econometric inputs to the price analysis NN to see if the they add predictive value (at least for longer timeframes).

In the meantime, if you're interested in the application of neural networks for econometric forecasting, you might check out the following paper:

http://www.smartquant.com/references...ks/neural6.pdf

For those of you that don't have time to read it, an interesting point they make is that forecasting the next couple months is very difficult, but the general long-term trend (6-12 months in the future) is much easier. Nothing we didn't really already know, but the rest of the paper has some interesting information for those who like the math part.

Wow, that references section in the website is quite fantabulastic. Seriously, these guys did a lot of reading.. almost every heavy quant topic has a list of references there, including things like option pricing, monte carlo..etc. And the NN section has papers I never even saw when I was doing my research! Thanks for the link!

I just wish I had time to read right now :'(
Virtue finds and chooses the mean. Aristotle, Ethica Nichomachea
  • #55
  • Last Post: Mar 12, 2008 1:24pm Mar 12, 2008 1:24pm
Quoting drummer486
I've decided to shift my focus away from news prediction (for now) and back to price analysis with NNs.
well good luck man. personally that NN stuff has always scared me (and im a programmer!). that is when you get into the big-boy leagues, competing against all these PHDs that the hedgefunds have hired. but dont let that worry you, the "little guy" has a way of sneaking in there and stealing profits. please keep us updated on your success, i dont think we have any successful NN traders at FF, youd be the first i believe!
Relax and be happy.
0 traders viewing now