Topic: Ross Hook
I use Ross Hook because I find it easy to get good Sharpe ratio numbers with it.
I am wondering if there are any adjustable parameters that can be applied to this indicator to make it fit more situations.
Create and Test Forex Strategies
You are not logged in. Please login or register.
Forex Software → Forex Strategy Builder (legacy) → Ross Hook
I use Ross Hook because I find it easy to get good Sharpe ratio numbers with it.
I am wondering if there are any adjustable parameters that can be applied to this indicator to make it fit more situations.
I think it's either a Ross Hook or it's not.
Actually this indicator does not represent Ross Hook very well. It doesn't look for passed 1-2-3 pattern.
It is not exactly Ross Hook but something like 4 bars fractal.
It does not accept parameters.
when FSB backtesting with Ross Hook as the opening point of the position with Take Profit as the Closing Point, I hope I'm wrong, but it seems to me that only those Ross Hook entries that result in the designated TP are counted. That is, there are no "losing trades". don't understand how OOS works in FSB if that is true. Weird.
when FSB backtesting with Ross Hook as the opening point of the position with Take Profit as the Closing Point, I hope I'm wrong, but it seems to me that only those Ross Hook entries that result in the designated TP are counted. That is, there are no "losing trades". don't understand how OOS works in FSB if that is true. Weird.
OOS - out of sample, it means defined part of data is cut off while optimizing/generating.
About your issue - your closing point is "take profit", what does it mean? It means that the position gets closed ONLY if it HITS the TP level. You see, therefore there can't be "losing trades".
when FSB backtesting with Ross Hook as the opening point of the position with Take Profit as the Closing Point, I hope I'm wrong, but it seems to me that only those Ross Hook entries that result in the designated TP are counted. That is, there are no "losing trades". don't understand how OOS works in FSB if that is true. Weird.
Is this a strategy that you built or generated...... generator is going to give you good results.
You can change the time period and remove OOS to test.... change the take profit and stop loss to test..
The nature of FSB is to show you constantly improving results per your settings for top 10, perhaps look at worst results.
I have often used this indicator in real time trading and had no problems.
can't figure out how to export to forum FSB Strategy Generator pix. Was using the v2.98.3 RC and had Ross Hook as "linked" opening slot and Take Profit as "linked" closing slot with 4 possible Opening Logic Conditions possible. When generating OOS=30 sometimes several other indicators would appear as opening logic conditions as a better strategy, but would almost always soon disappear, yielding to better strategies not involving them. If all disappeared, then there would be a very good sharpe and no 'Losing" trades. If other conditions didn't disappear then there would still be about 2 wins per loss. If Ross Hook has no adjustable parameters I don't understand why further generation would cause the other conditions to disappear. the only other parameter that can be adjusted is the Take Profit. Seems there must be a logic error in FSB Ross Hook that basically doesn't count Ross Hook entries that reverse and show a loss.
About your issue - your closing point is "take profit", what does it mean? It means that the position gets closed ONLY if it HITS the TP level. You see, therefore there can't be "losing trades".
How about when I use a "linked" Trailing Stop Limit. I still get virtually no "losing trades" even in the OOS regions.
About your issue - your closing point is "take profit", what does it mean? It means that the position gets closed ONLY if it HITS the TP level. You see, therefore there can't be "losing trades".
How about when I use a "linked" Trailing Stop Limit. I still get virtually no "losing trades" even in the OOS regions.
Do you still detect unusual behaviour when slot is "locked"?
OOS is independent of the data that you are working with, somewhere in the forum Popov explained that.
I understand your ques re OOS, I think Footon has explained.
I do not think there is an error, I have used this indicator extensively and I am guessing that Footon has examined it on the inside for his own purposes.
You could open the chart and examine the entries to detect potential errors......
Popov will be pretty quick to fix any error that you may detect.
As I understood OOS, the strategy generated did not depend on the data in it. Once a strategy has been generated, then it is applied to the OOS data. that is why a strategy that works on the OOS data just as well is especially to be valued. Presumably many of the Ross-StopLimit generated strategies will not do well at all on the OOS. But mine do. There's something wrong. Many, perhaps most of such entries in the real world will result in losses.
The beauty of FSB is that you can change parameters and generate as many strategies as you want and demo them.
Certainly you will want to demo a few of your strategies to prove or disprove your theory.
Please do report your findings as Popov will want to make adjustments or corrections.
Post a couple of your strategies and Footon and I will have look.
As I understood OOS, the strategy generated did not depend on the data in it. Once a strategy has been generated, then it is applied to the OOS data. that is why a strategy that works on the OOS data just as well is especially to be valued. Presumably many of the Ross-StopLimit generated strategies will not do well at all on the OOS. But mine do. There's something wrong. Many, perhaps most of such entries in the real world will result in losses.
First you say Take Profit, now I read Stop Limit, bloody hell, which one is it? And post those strats for a check-up, like Dave said.
don't know how to post them. Neither Win7 nor Win8 has a "clipboard" you can save files to, "look at", choose a file saved there and transfer to another location, or attach to an email, whatever. There are allegedly third-party apps you can buy that will allegedly allow you to recover a useful 'clipboard'. And I no longer use WordPerfect, Excel, usw. Meanwhile I have redone all those Ross-TakeProfit linked 'strategies' that delivered no losing trades, making them Ross-TrailingStopLimit 'linked' strategies that deliver almost no losing trades, even for the OOS-30 data. If it's "either a Ross indicator or it isn't", looks to me like it isn't.
jgp wrote:when FSB backtesting with Ross Hook as the opening point of the position with Take Profit as the Closing Point, I hope I'm wrong, but it seems to me that only those Ross Hook entries that result in the designated TP are counted. That is, there are no "losing trades". don't understand how OOS works in FSB if that is true. Weird.
OOS - out of sample, it means defined part of data is cut off while optimizing/generating.
About your issue - your closing point is "take profit", what does it mean? It means that the position gets closed ONLY if it HITS the TP level. You see, therefore there can't be "losing trades".
Hi Footon,
you should be right but you are not. Pls let me explain. If you use FSB to generate strategy that would
A. add to position
B. has TP as closing logic
this would result in trades that would
A. Hit TP
B. and produce the loss at the same time. yes right, position will hit TP but the result of positions basket is loss.
you should be right but you are not.
Picture says more than 1000 words, right.
What I said is correct! You alter my statement by adding another element to the equation and say I'm wrong... well, I'm speechless.
To J, do you have a menu bar in your FSB? Maybe you hit Strategy and then Publish, and then copy/paste your holy grail here?
I ran Ross Hook strategies for three hours last night, all sorts of them........ good ones and bad ones and mean ones and nice ones looking for what you are describing. No luck.
There is, at the top of FSB a menu 'Strategy' and inside that menu is 'Publish'. That little item has nothing to do with Microsoft , in fact it is something that FST has developed for this very purpose.
Use that thing to post your strategy.
This FST while it can do myriad things is not complex to operate, a person can keep it really simple and get excellent results. Footon has spent the time to learn a lot about indicators and the language, he can probably analyze things very quickly...... that is if you convey the message simply.
As I said earlier, post the strategy and we will do our best and if there is some error, it will get fixed forthwith, however, keep it simple.
Strategy name: Generated
Forex Strategy Builder v2.98.3.0 RC
Exported on: 6/15/2013 10:09:38 AM
Description:
(This description might be outdated!)
Automatically generated on 6/14/2013 7:08 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 51885.40 USD (8/29/2012 4:00 AM Bar: 2591)
-----------
Modified by the strategy generator on 6/14/2013 7:45 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 52680.10 USD (8/29/2012 10:00 PM Bar: 10316)
-----------
Modified by the strategy generator on 6/14/2013 8:35 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 55632.10 USD (8/29/2012 10:00 PM Bar: 10316)
-----------
Modified by the strategy generator on 6/14/2013 8:42 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 56038.80 USD (8/29/2012 9:00 PM Bar: 10317)
-----------
Modified by the strategy generator on 6/14/2013 9:07 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 50626.10 USD (11/21/2012 3:30 AM Bar: 13999)
-----------
Modified by the strategy generator on 6/14/2013 9:16 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 51787.60 USD (11/21/2012 3:30 AM Bar: 13999)
-----------
Modified by the strategy generator on 6/14/2013 9:21 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 51787.60 USD (11/21/2012 3:30 AM Bar: 13999)
-----------
Modified by the strategy generator on 6/14/2013 9:35 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 59218.20 USD (8/29/2012 4:00 AM Bar: 2591)
-----------
Modified by the strategy generator on 6/15/2013 10:07 AM.
Market: EURUSD 4 Hours
Spread in points: 20.00
Swap Long in Points: 2.00
Swap Short in Points: -2.00
Commission per lot at opening and closing in Points: 0.00
Slippage in points: 0
Maximum open lots: 1.00
Entry lots: 0.10
Adding lots: 0.10
Intrabar scanning: Accomplished
Interpolation method: Pessimistic scenario
Ambiguous bars: 1
Tested bars: 18529
Balance: 4153 points (54153.00 USD)
Minimum account: -1 points (49998.80 USD)
Maximum drawdown: 1028 points (1028.40 USD)
Time in position: 19 %
[Strategy Properties]
A same direction signal - Adds to a winning position
An opposite direction signal - Closes the position
Permanent Stop Loss - None
Permanent Take Profit - 600
Break Even - 2100
[Opening Point of the Position]
Ross Hook
Enter long at an Up Ross hook
[Opening Logic Condition]
Top Bottom Price
The bar opens above the bottom price
Base price - High & Low
Base period - Previous week
Vertical shift - 371
[Opening Logic Condition]
Trix Index
Trix Index line is higher than zero
Smoothing method - Simple
Base price - Weighted
Period of Trix - 116
Use previous bar value - Yes
[Opening Logic Condition]
Accelerator Oscillator
AC falls
Smoothing method - Weighted
Base price - Close
Slow MA period - 140
Fast MA period - 64
Forming MA period - 90
Level - 0.0000
Use previous bar value - Yes
[Opening Logic Condition]
Day of Month
Enter the market between the specified days
From (incl.) - 30
Until (excl.) - 16
[Closing Point of the Position]
Take Profit
Exit at the Take Profit level
Take Profit - 266
If I continue to generate this Ross-TP strategy, most or all of the opening logic conditions will disappear.
What is the source of the data I do not have that many bars so I want to get the same data. Is your data clean, ie are all the bars there or are there missing bars?
you have a TP of 266 and a BE of 2100 and a permanent TP of 600, can you align the two TP numbers and get a proportionate BE. There is no stop loss.
Go to 'Testing' on the menu bar and click on 'Additional Statistics' so we get some facts in addition to what you posted. What have you got the Sharpe Ratio set to...... and what have you got the 'Maximum Equity Drawdown' set to?
Until I understand your data I can not duplicate the situation.
Please let me know the data source and change those parameters. Then repost the strategy
I suggest strongly that you incorporate a stop loss or the unpleasant things will occur for sure.
here is a RossHook-TakeProfit linked strategy being generated on latest FSB RC
trategy name: Generated
Forex Strategy Builder v2.98.3.0 RC
Exported on: 6/15/2013 11:51:33 AM
Description:
(This description might be outdated!)
Automatically generated on 6/14/2013 7:08 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 51885.40 USD (8/29/2012 4:00 AM Bar: 2591)
-----------
Modified by the strategy generator on 6/14/2013 7:45 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 52680.10 USD (8/29/2012 10:00 PM Bar: 10316)
-----------
Modified by the strategy generator on 6/14/2013 8:35 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 55632.10 USD (8/29/2012 10:00 PM Bar: 10316)
-----------
Modified by the strategy generator on 6/14/2013 8:42 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 56038.80 USD (8/29/2012 9:00 PM Bar: 10317)
-----------
Modified by the strategy generator on 6/14/2013 9:07 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 50626.10 USD (11/21/2012 3:30 AM Bar: 13999)
-----------
Modified by the strategy generator on 6/14/2013 9:16 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 51787.60 USD (11/21/2012 3:30 AM Bar: 13999)
-----------
Modified by the strategy generator on 6/14/2013 9:21 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 51787.60 USD (11/21/2012 3:30 AM Bar: 13999)
-----------
Modified by the strategy generator on 6/14/2013 9:35 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 59218.20 USD (8/29/2012 4:00 AM Bar: 2591)
-----------
Modified by the strategy generator on 6/15/2013 11:47 AM.
-----------
Modified by the strategy generator on 6/15/2013 11:48 AM.
Market: GBPUSD 4 Hours
Spread in points: 20.00
Swap Long in Points: 2.00
Swap Short in Points: -2.00
Commission per lot at opening and closing in Points: 0.00
Slippage in points: 0
Maximum open lots: 1.00
Entry lots: 0.10
Intrabar scanning: Accomplished
Interpolation method: Pessimistic scenario
Ambiguous bars: 0
Tested bars: 18706
Balance: 3712 points (53711.60 USD)
Minimum account: 0 points (50000.00 USD)
Maximum drawdown: 0 points (0.00 USD)
Time in position: 11 %
[Strategy Properties]
A same direction signal - Does nothing
An opposite direction signal - Does nothing
Permanent Stop Loss - None
Permanent Take Profit - None
Break Even - None
[Opening Point of the Position]
Ross Hook
Enter long at an Up Ross hook
[Opening Logic Condition]
Bar Range
Bar Range falls
Number of bars - 59
Level - 0
Use previous bar value - Yes
[Opening Logic Condition]
Oscillator of CCI
Oscillator crosses the zero line upward
Smoothing method - Exponential
Base price - Typical
First CCI period - 28
Second CCI period - 13
Use previous bar value - Yes
[Closing Point of the Position]
Take Profit
Exit at the Take Profit level
Take Profit - 470
This strategy is being generated by FSB v2.98.3.0 RC on data obtained from AlpariUS charts downloaded onto a FlashDrive as was suggested by Master Popov when Win8 was refusing to allow me to access that same data when downloaded to FSB/Users/Data.
I think you are going to have to set up some criteria before you start the generator, without a stop loss you will burn.
by the way, I began all RH-TP linked generations with NothingNone strategy properties with no selected additional Opening Logic conditions but with the possibility of 4-5 being added. here is the same generation [all of the top ten generated strategies have zero losses and zero drawdowns and nice high sharpe values] posted above
Strategy name: Generated
Forex Strategy Builder v2.98.3.0 RC
Exported on: 6/15/2013 12:23:10 PM
Description:
(This description might be outdated!)
Automatically generated on 6/14/2013 7:08 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 51885.40 USD (8/29/2012 4:00 AM Bar: 2591)
-----------
Modified by the strategy generator on 6/14/2013 7:45 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 52680.10 USD (8/29/2012 10:00 PM Bar: 10316)
-----------
Modified by the strategy generator on 6/14/2013 8:35 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 55632.10 USD (8/29/2012 10:00 PM Bar: 10316)
-----------
Modified by the strategy generator on 6/14/2013 8:42 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 56038.80 USD (8/29/2012 9:00 PM Bar: 10317)
-----------
Modified by the strategy generator on 6/14/2013 9:07 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 50626.10 USD (11/21/2012 3:30 AM Bar: 13999)
-----------
Modified by the strategy generator on 6/14/2013 9:16 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 51787.60 USD (11/21/2012 3:30 AM Bar: 13999)
-----------
Modified by the strategy generator on 6/14/2013 9:21 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 51787.60 USD (11/21/2012 3:30 AM Bar: 13999)
-----------
Modified by the strategy generator on 6/14/2013 9:35 AM.
Out of sample testing, percent of OOS bars: 30%
Balance: 59218.20 USD (8/29/2012 4:00 AM Bar: 2591)
-----------
Modified by the strategy generator on 6/15/2013 11:47 AM.
-----------
Modified by the strategy generator on 6/15/2013 11:48 AM.
-----------
Modified by the strategy generator on 6/15/2013 12:08 PM.
Market: GBPUSD 4 Hours
Spread in points: 20.00
Swap Long in Points: 2.00
Swap Short in Points: -2.00
Commission per lot at opening and closing in Points: 0.00
Slippage in points: 0
Maximum open lots: 1.00
Entry lots: 0.10
Intrabar scanning: Accomplished
Interpolation method: Pessimistic scenario
Ambiguous bars: 0
Tested bars: 18714
Balance: 3944 points (53944.40 USD)
Minimum account: 0 points (50000.00 USD)
Maximum drawdown: 0 points (0.00 USD)
Time in position: 10 %
[Strategy Properties]
A same direction signal - Does nothing
An opposite direction signal - Does nothing
Permanent Stop Loss - None
Permanent Take Profit - None
Break Even - None
[Opening Point of the Position]
Ross Hook
Enter long at an Up Ross hook
[Opening Logic Condition]
Bar Range
Bar Range falls
Number of bars - 51
Level - 0
Use previous bar value - Yes
[Opening Logic Condition]
Oscillator of CCI
Oscillator crosses the zero line upward
Smoothing method - Exponential
Base price - Typical
First CCI period - 28
Second CCI period - 13
Use previous bar value - Yes
[Closing Point of the Position]
Take Profit
Exit at the Take Profit level
Take Profit - 470
Please refer to my message above and look after those questions and suggestions......
Using 4 or 5 empty slots is not really productive, use 2 or 3 at the most, preferably 2 and ensure that you set some criteria such as stop loss and sharpe ratio so that you get some worthwhile results.
When generating and testing, probably preferable to use a fixed lot size, no adding or subtracting from the position as results will be skewed.
Thanks for your support thus far. My sense is that the data source and quality is essentially irrelevant to the weird strategies generated. My concern is that for some reason this FSB version is frequently generating strategies with wonderful statistics and outstanding balance/equity charts that are literally too good to be true. You said yourself that you liked Popov's Ross Hook because it gave such excellent Sharpe values. Why does it?
the generator changes my NothingNone settings to whatever; I don't. and I always have my minimum Sharpe setting= 0.2, always.
No offense intended, but my guess is that somewhere in this latest FSB's Generator there's a Phantom Fudge Factor that mysteriously takes on whatever value it has to, to make the generated strategies look very much better than they really are.
Forex Software → Forex Strategy Builder (legacy) → Ross Hook
Powered by PunBB, supported by Informer Technologies, Inc.