PDA

View Full Version : Sixty (60) Back Tests w/Settings & Results In Attached Excel Workbook


rdalfrey
11-30-2008, 08:58 PM
First off a belated Happy Thanksgiving to each of you and your family.

I had hoped I could have finished all these back tests and posted them before this weekend so each of you would have had time to review the results before the market opens today.

I started running these back tests last weekend and needless to say I am plumb wore out on back tests. My wife after 45 years of marriage to me threaten to divorce me if I didn't stop and pay more attention to her. Fellows, this can be very addictive so be very careful.

I apologise for the length of this post but please bear with me.

I would hope that you print out this post (click file, print preview, adjust size to 100 or 125%, and click the print tab) and the attached Excel Workbook (only the first nine pages under Russell's Back Test Tab), go sit in a corner somewhere with a Scotch & Water, and review my observations of the back tests I performed.

By the way, I configured the Excel Workbook whereby you have blank pages under the three "Your Back Tests" tabs at the bottom of the Excel Workbook for you to use and record your Back Tests if you want to.

It sure makes it easier to compare your Back Tests when you have the results side by side.

Please feel free to critique my observations below and maybe you will also find additional worthwhile info that I didn't see that could be beneficial to us.

My observations:

BACK TEST NUMBER 1 THRU 9

I know M1 modeling quality is not that good and the short time frame, that's all the data I could get from IBFX, is not conclusive but all I was interested in was to find out what was the best Take Profit, LRR, and Max Orders settings for a small Mini account.

My conclusions:
1. Take Profit of 25 is the best.
2. LRR of 2 is the max the acct can stand with a T/P of 25. (See Test #9).
3. Max Order of 1 is the best. Higher than 1 will wipe out acct. (See Test #8) If you want more return increase your LRR, not your Max Orders.

BACK TEST NUMBER 10 THRU 18

These tests were performed to find out the best Moving Average settings with different aaa & bbb settings.

My conclusions:
1. Default aaa, bbb, and MA was the best.
2. Different MAs with same aaa & bbb setting had no effect on Net Profit.
3. Note: Different MAs does have an effect on M5 but not on M1.

BACK TEST NUMBER 19 THRU 27

These tests were performed on M5 to find out the best aaa & bbb settings with different Moving Averages.

My conclusions:
1. aaa=30 & bbb=15 were the best w/all MAs. (See test #21, #24 & #27)
2. FAPS default aaa & bbb settings does not like M5. (See #19, #22 & #25)
3. Different EMAs did not effect Net Profit. (See #23 & #24 and #26 & #27)

BACK TEST NUMBER 28, 31 & 34

I performed these tests with the same settings as test #19, #22, & #25 respectively to see if reducing the LRR would have an effect.

My conclusions:
1. It definitely does.
2. Proves money management is essential to keep from wiping out acct.

BACK TEST NUMBER 37 THRU 45

As you will note these tests were performed on my Forex Meta Demo account with a longer time frame for a different perspective.

The tests were performed with different aaa & bbb settings for the same Moving Average.

My conclusions:
1. Again FAPS default aaa & bbb does like M5. (See text #37, #40 & #43)
2. aaa=30 & bbb=15 was the best. (See test #39)
3. Default MA of 100 was the best. (See test #39)

BACK TEST NUMBER 46, 49 & 52

Again I performed these tests with the same settings as test #37, #40, and #43 respectively to see if a LRR cut in half would effect the Net Profits.

My conclusions:
1. Again it definitely does. From a negative to positive Net Profit.
2. Again Money Management is essential for survival of your account.
3. You can not push the envelope just to be greedy. Patience is vital.

BACK TEST NUMBER 55 THRU 64

These tests were performed to see if using a Stop Loss with the best settings so far would improve or hinder Net Profits.

My conclusions:
1. It definitely does.
2. The lower your Stop Loss the lower your Net Profit except for test #62
3. FAPS does not have an internal Stop Loss as advertised.
4. The only Stop Loss FAPS has is close all orders at the End of the Month.
5. See test #64 where a Stop Loss of 1000 was needed to equal test #39
6. Proof: FAPS closed EOM order on 11/3 with a 910 PIP loss.

BACK TEST NUMBER 73 THRU 81

I performed these tests, sorry on a short time frame, just to see how different aaa & bbb settings would effect Net Profits. As you can see aaa=30 and bbb=15 provided the best Net Profit again. Maybe there is more tweaking that could be done with these settings but I am not up to trying them right now - I don't want to see those divorce papers as I can't afford it. HA!

CLOSE AT EOM (End of Month)

This function of FAPS is very alarming to me. Take a look at the Close At EOM "$ Amount" on each test that I ran and tell me if that doesn't concern you. Granted FAPS still ended up with a good Net Profit but what would it have been if these orders weren't closed. Out of all the orders that were closed at EOM there were only 23 of 186 that were positive closes at under 25 PIP profit. I checked on some of the orders that were closed and they would have turned into a profit in 2 to 5 days if they had not been closed. On the other hand, like the one closed on 11/3 at a 910 PIP loss, needed to be closed as the trend was definitely against the trade. As a matter of fact the trade would still be opened today with an additional loss of 100 to 150 PIPS.

Does anybody know why FAPS closes all orders at EOM? Why not close all orders in the middle of the month? Why not close all orders twice a month? I would venture to say that FAPS does not have the necessary coding to realize it has made a trade that is bucking the trend and thus must rely on the market coming back or close at EOM to get out of it's bad trade. Any thoughts on this?

Maybe our coders on the forum can help in this area.

My plans to defeat this FAPS function is to disable FAPS before midnight, platform time which would be 6PM CST on IBFX, on the last trading day of the month if the trade is negative and the trend is going sideways or maybe trying to change. I will monitor the trade after midnight and decide whether to let it run or manually close it. I will not turn FAPS back on, if I decide to let the trade run, until the trade closes with a profit or I manually close the trade.

The reason I will not turn FAPS back on after midnight is because I do not know if FAPS will close that trade regardless of when you turn it back on after midnight.

I will know FAPS reaction to my plans tomorrow as I had an open trade on my Demo account and I disabled FAPS and the trade is still running in a negative trend. I will turn FAPS back on tomorrow and see what it does with that trade and I will let everyone know the results.

In case you are wondering where I got these Close at EOM figures, I tallied them up from the Results Tab of the Tester as I ran each test.

Another observation while running all of these tests - if FAPS had a bad month it did not make a difference what settings were used. All settings had likewise results at EOM close. Percentage wise I would assume a little different but not much.

One final thought on something else before I close. Next Friday @ 8:30AM EST, I think, the NFP (Non Farm Payroll) Report will be released and it is expected to show a great increase in unemployment which will cause a volatile reaction in the Forex market.

Regardless of what the report says I plan to try and have FAPS turned off with no open trades before the release and up to 30-45 minutes after the release of the report.

That's it for now.

Your comments would be appreciated.

Russell

oofdah
11-30-2008, 09:18 PM
Excellent work Rdalfrey!!! Many Kudos!!

I too have been doing something of this sort, however my results are handwritten. Just about 60 last count.

I will compare your tests with mine and see if we can do even better ;)

rdalfrey
11-30-2008, 09:48 PM
Thanks

Don't forget you can record your settings and results in "Your Test" tabs at the bottom of the WorkBook.

Let us know your results. You could attach the same Excel WorkBook with your results for us to see in one of the "Your Back Tests" tabs.

Just rename the tab by right clicking on it, choose rename, move your cursor into the tab and type in the name you want, and press enter.

I'm off of back testing for a while. It has completely wore me out.

Russell

oofdah
11-30-2008, 09:54 PM
Have a beer, on me ;)

adamshea
11-30-2008, 10:12 PM
so overall what were all over your best settings (in one message to make it easier to cypher through

rdalfrey
11-30-2008, 10:47 PM
so overall what were all over your best settings (in one message to make it easier to cypher through

Evidently you did not read the entire post much less compare my conclusions to the back tests settings and results on the Excel Workbook.

The purpose of all the Back Tests I did was for each person to review them and decypher themselves what is the best settings. That is how we learn.

My conclusions might be correct but then again it might not be. But if you want the settings without digging in yourself, read my conclusions and look at the results in the Excel Workbook.

adamshea
12-01-2008, 02:19 AM
my bad, didn't see the excel download

rdalfrey
12-02-2008, 12:35 PM
As promised, you can defeat FAPS' function of closing any open trades at the End of the Month.

On my Demo acct I had an open trade that was opened on 11/28 and was down approx 110 Pips.

I turned FAPS off at approx 30 minutes before midnight, platform time, and turned it back on at the start of 12/2. The trade was still down.

FAPS did not close the trade and let it continue running.

The trade finally closed, with a 25 Pip profit, approx 3 hours ago. Instead of 110 Pip loss that trade generated a 25 Pip profit. The only question that remains is will it work everytime and how many profit trades could FAPS have generated in that time span.

I do know it would have taken 5 winning trades in that time span to make up for the EOM closure if I would have let FAPS do it's thing.

So, if you so desire, you can stop the EOM closures by turning FAPS off and then turning it back on after the new month starts. How soon I don't know but I know waiting one day defeated it's EOM closure function.

Russell

AddictedNewbie
12-03-2008, 05:37 PM
Hello there, are the results based on 90% modeling quality on the timeframes tested?

Thanks!

rdalfrey
12-03-2008, 11:06 PM
Hello there, are the results based on 90% modeling quality on the timeframes tested?

Thanks!

Yes, they are except for the M1 runs which were run only to establish certain criterias as indicated in my conclusions for the M5 runs.

Russell