Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

How to Correctly BackTest, Collect Data, and Analyze A System 13 replies

How much historical data needed to backtest day trading system 0 replies

Visual backtest result different from non-visual backtest 0 replies

data needed to backtest system 1 reply

How to backtest system with past data? 8 replies

Attachments: NFP System - Backtest data provided
Exit Attachments

NFP System - Backtest data provided

Quoting billbss
There have been some significant gaps in price (in the opposite direction from the initial spike) that scare me away from NFP trading.
Hi billbss, sure it depends on the broker and on settings. Test it with a straddle (position in both directions) with a pending order 5-2min before the news comes out. Pend order gap should be at least 25Pips to both sides according to the current price you have. Set your S/L to 25 and your T/P to 25. I trade it primarely on EUR/USD, but you can use it on GBP/USD as well (take everything to 30 then). Once you are in profit, don't let it go against you, even go for the spread. And if you can, test it live with Micro-Lots. I haven't lost one trade since I use this (on NFP & FOMC, others are more difficult) manually.

Quoting Kevin_E
Matrixler, thank you for your insight and your kind words I'll look into these platforms and do some tests when the next NFP comes around. If indeed orders such as this are possible, then I wouldn't have any quams integrating the two systems at all. I think it would turn profits much quicker. Always have to keep an open mind.

Also, I'm making some minor to the system. No rule will be added to create new orders or filter out orders, so the system will remain the same. The only thing that I am going to change is the system of stop losses and take profit. I put this system together with a limited amount of time for a contest, and as satisfied as I am with it, I think it can be improved on quite a bit. A reversal system will probably be set in place to limit losses, and a better take profit system will be put together for those who hate seeing your price retreat 150 pips before it hits the 336 SMA to close. Of course, as I make new discoveries, you at FF will be the first to know

-Kev
Savage Pip Fiend
Thanks Kevin, for keeping an open mind and give it a try to test.
It's always hard to optimize a system, even if it's for a long term strategy, because of the fact that one has to wait for a longer time to see it's impact on the current market conditions.
But if your backtests show you any improvement to the risk/reward ratio and max drawdown then you should defenitely give it a "try".
Again, I like the system and your way of sharing it with us.

Maybe we found some cool programmer here who is able to code this one for MT4.

Good trades to all of you
Has had EA for NFP???
Warm Regard { Links removed }
Hi Mr. Winner, PM me for that.

Quoting the winner
Has had EA for NFP???
hi Matrixler

where i can see the rules of nfp system thet you mention.

thanking you in advance

btfl
hey kevin, I sent you a pm, please check it. thanks.
Contact me for Custom Indicators/EAs.
Matrixler, I was talking to kevin. I dont know who you are...are you related to Kevin?
Contact me for Custom Indicators/EAs.
Quoting Kevin_E
Also, I'm making some minor to the system. No rule will be added to create new orders or filter out orders, so the system will remain the same. The only thing that I am going to change is the system of stop losses and take profit. I put this system together with a limited amount of time for a contest, and as satisfied as I am with it, I think it can be improved on quite a bit. A reversal system will probably be set in place to limit losses, and a better take profit system will be put together for those who hate seeing your price retreat 150 pips before it hits the 336 SMA to close. Of course, as I make new discoveries, you at FF will be the first to know

-Kev
Savage Pip Fiend
Kevin, how about something simple to start with, such as adding a 200EMA or something that will catch the price before it turns sooner? Of course this will also make it more prone to false close signals, but it might be worth looking into...?
Quoting zalakinc
Matrixler, I was talking to kevin. I dont know who you are...are you related to Kevin?
Hi Zalakinc, no I'm not. On my way of searching through the net about NFP/News trading, I found this.
I'm from Germany, trading forex for 1 Year now and like News Trading especially the NFP. Kevin's System is a perfect addition for me, because of the long term strategy.
To implement the other strategy was just a suggestion.
Quoting VLA
Hi friends, I met the test GBPUSD of history in 2006 and received these results :
January +100
February +7
March -100
April and May +1049
June +303
July +200
August +8
September +295
October +305
November -21
December -400
Thanks VLA for your post.

I've compared your back testing to Kevin's below. It's not clear if Kevin backed tested Dec but I've included your Dec result for worse case comparison.

Month VLA Kevin

January +100 +100
February +7 +58
March -100 +50
April and May +1049 +43 (Apr=+317 May=-274)
June +303 +193
July +200 +50
August +8 (No trade)
September +295 +150
October +305 +305
November -21 +393
December -400 (No trade??)

Total VLA=1746 Kevin=1342


You did very well! Congrats! Any idea why your results were different to Kevin's (esp. Apr/May)?

Has anyone else done any independent backtesting? I'd be interested to see your results!
Because this is a long term strategy, a swap free broker would be ideal. Does anyone have one they can recommend? Thanks.
Sorry for not being around again for a few days. This is the last week before the next NFP, so I'm crunched for time on a number of projects.

The independent backtests should be ideally close to the same, but there are some very important factors to point out. When I did my backtesting, I was doing it fully manual on the FXCM feed in FXCM's TradeStation. All of my times were EST, and the Sentinel number I used had not been put into an mql4 custom indicator yet, so it was all hand updated. I assume many of you are doing the test in MetaTrader with the automated Sentinel Index, which could account for slight differences before you even get down to the testing part. Backtest it how YOU would like to trade it, and optimize according to your schedule and preferences. I made this system public so people could take what I had done, and expand it to fit their own method of trading. If you achieved outstanding backtest results from my system, then by all means, trade it how you tested it

I'm very interested in additions people have made or would like to see to the system, and look forward to building something even better together than what I have made alone.

JoshDance, I have looked into different MA values when coming up with the take profit scenario, and none (not even the 336 that I seem to keep using) seem to be the most efficient in catching a move and limiting losses. Your price is either going to turn around for 100+ pips until it hits your larger MA, or will tag your smaller MA much sooner than it should and stop your move out completely. I've come up with a new indicator called the Storm number that'll remedy this, and it's based off of the slope and movement of a specific price MA. I haven't done any testing with this number against the NFP system, but when I do, you at FF will be the first to hear about it.

All right, I'm off to work a little more. I'm not taking signup requests for the month of PDF signals anymore, as this Friday is our next NFP release, and I feel a re-entry trade this week could prove risky. If you have sent a request within the past few days to be on the mailing list, I will be holding your email address until this Friday, and will send you all of the updates and signals in a zip file to your email for review. Until I can get them all uploaded on a web page, anybody can send me a PM and request the previous PDF files, and you should have them the same day. Of course, as always, free of charge Happy trading!

-Kev
Savage Pip Fiend
Kevin/others,

I heard on another thread that Monday will be a holiday for parts of europe? Is this true? If so, will this impact our NFP initial box timeframe?
Greetings. I have read through the thread and i understand the concept of what is being attempted and i like the idea as it meshes with my compulsion to breakout trading. One the difference between the three indicators in post 1 that have been updated a few days ago i have a question.

All seems to be opperating fine, and i'm wondering which of the three people are finding more visually helpful?

If i'm interpreting the data correctly, as of friday's (may 4th) close on the GbpUsd 30m chart we have the price action closing about 20 pips below the 336 SMA. Now looking at the Sentinal Index which i've placed a 336sma set to first indicator i see the data line above the sma. I interpret this as the USD is in positive territory related to the other currencies.

Now onto the Sentinal Index_SMA indicator is visualy showing the inverse. I.E. as the USD goes up in value against the others the SI data line goes down very much like the GbpUsd price action might go down when the USD increases in value over the Gbp.

Both are working, but the first form, I.E. USD Bullish = Up / Bearish = Down makes more sence to me just like one might read the Williams%R indicator. What do those who have been using these indicators think, as i've just placed them on my chart for the first time today after digesting the thread's data.

Cheers,
Thom
Per a discussion with Kevin last night, on his general advice I extended the box period to go till the start of the NY session this morning at 8AM ET due to the UK holiday today and low market activity. That increased the upper limit of the box by a few pips, but nothing major.

Currently, my no-trading zone for GU is:

High: 1.9974 (hit 07:30GMT on 7/5/07)
Low: 1.9846 (hit 13:30GMT on 4/5/07)

(data from Alpari UK)

So, right now no trades. But these are the numbers I'll use for my trading this month, unless someone else thinks that we should extend the box further...
oh ok. Good thing it doesnt make a whole lot of difference in this case. I am alreayd in a trade at 1.9970 according to the original rules fo the system. Already down about 40 pips. I hope it works out well.
Contact me for Custom Indicators/EAs.
Quoting zalakinc
oh ok. Good thing it doesnt make a whole lot of difference in this case. I am alreayd in a trade at 1.9970 according to the original rules fo the system. Already down about 40 pips. I hope it works out well.
Well, hopefully that will turn up for you... I'd rather miss out on a few pips up front and be a little more sure of the direction, so even if a candle closed above/below the zone an hour or so after the no-trade-zone ended, I'm still waiting at least a day to try and avoid losing pips upfront.
Quoting JoshDance
Well, hopefully that will turn up for you... I'd rather miss out on a few pips up front and be a little more sure of the direction, so even if a candle closed above/below the zone an hour or so after the no-trade-zone ended, I'm still waiting at least a day to try and avoid losing pips upfront.
hi :

Can someone please post a datailed live statement ?

Thanks
As of now we could have technically taken our first long trade for the month. Price closed on 30M above the 1.9974 level. However, it's consolidating a little now, and the sentinel is as of right now back a bit below its MA. Keep your eyes open though!
We gave it two chances this month and it looks like it blew them both.
I know, well the second chance hasnt been blown yet....and even if it does,2 trades is not a big deal. it kind of worries me a bit though wiht its 80% accuracy published.
Contact me for Custom Indicators/EAs.
0 traders viewing now