stevehopwoodforex.com
https://www.stevehopwoodforex.com/phpBB3/
Print view

Bob and Shelley Again
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=2759		 Page 78 of 120
Author:  fxozgirl [ Fri Nov 22, 2013 10:51 pm ]
Post subject:  Re: Bob and Shelley Again
Look at the trades posted...the mix of trades are different for everyone! Different brokers, different feeds, different trades and that goes between demo & live as well

Also depends on what settings you are using as well.
Author:  SteveHopwood [ Fri Nov 22, 2013 11:54 pm ]
Post subject:  Re: Bob and Shelley Again
fxozgirl wrote:Look at the trades posted...the mix of trades are different for everyone! Different brokers, different feeds, different trades and that goes between demo & live as well

Also depends on what settings you are using as well.
It is even worse than that.

I have BASA on my live Cowboy IBFX account, also on a live Cowboy IBFX account that I run for a much-loved former student. Both Cowboy IBFX Aussie micro accounts trading the same charts with the same lot sizes. Both accounts lost a bit this week but I was able to close all my trades at a profit before close of business. I have left Gareth's open because his basket never reached profitability.

So, we gave the same criminal + 2 live accounts + same bot trading same lot sizes on the same charts = different results. Wow.

--------------------------------------------------------
This week both my accounts in positive pips but cash loss. What is causing this is trades that I allow to go to full stop loss. Going there involves crossing back over the 240 in the wrong direction, so closing them as soon as this happens is clearly a good idea.

I have moved entirely to the H1 chart for trading BASA. Here is a pic of a big-losing UC trade. The moving average are converted to the H1, so the 240 H4 becomes the 920 H1. Andy's stacking ma's removed to get rid of some clutter.
usdchfh1.png
It is a picture perfect trade as presented. Cutting the trade off following a pass of the H4 240 would have saved a fair few pips. Maybe closing it at the half way point between the two ma's?

But now look at the H4 pulled out:
usdchfh4.png
Anyone in their right minds still contemplating buying UC? Daily is a bit more mixed until you look at the weekly. :lol:

-------------------------------------------------------------------------
Then there is this: I have seen BASA this week correctly detecting an uptrend (60 > 240) but buying when the market is below the 240. There must be a programming fault in there somewhere because BASA should only ever buy when Bid > 60 > 240 or whatever your calculated equivalents are on lower than H4 tf charts are).

Bugger.

:D .
Author:  matrixebiz [ Fri Nov 22, 2013 11:56 pm ]
Post subject:  Re: Bob and Shelley Again
fxozgirl wrote:Look at the trades posted...the mix of trades are different for everyone! Different brokers, different feeds, different trades and that goes between demo & live as well

Also depends on what settings you are using as well.
Ok, thanks for the confirmation. I think I'll have to start a Live account soon :)
Author:  matrixebiz [ Sat Nov 23, 2013 12:48 am ]
Post subject:  Re: Bob and Shelley Again
SteveHopwood wrote:
fxozgirl wrote: This week both my accounts in positive pips but cash loss. What is causing this is trades that I allow to go to full stop loss.
True, we don't want trades to get to full stop loss if we can detect a bad entry early enough and get out of it before it goes too bad but those using the Close Half Lots function may also be cheating themselves out of Cash. I realized this when looking at the History indicator and I was in positive pips but negative cash and I was like 'what' how can that be, until I diagnosed it a little. I'll give a little example below for those that don't understand what I'm talking about.

Lets say I'm trading 20 cents a pip;

Scenario 1 (w/Half lot close) - I've just had a SL of -45 pips then another trade closed half lots at +30 pips and the other half closed at an additional +20 pips. So what just happened is the bad trade I lost 45 pips but positive 80 (30+50) pips in the good trade I closed Half to lock in profit, I'm ahead positive 35 pips, Oh yeah! .... but in reality when you look at the cash, -45 @ 0.2 = $9 and +80 @ 0.1 = $8, so I've actually lost $1

Scenario 2 (No half lot close) - I've just had a SL of -45 pips and if I can get a full trade to close at least +50 pips and not worrying about a +80 like above, I've only gained +5 pips profit but in cash I'm actually up $1 -45 @ 0.2 = $9 and +50 @ 0.2 = $10
So also, if we were in the exact same scenario this time as scenario 1 and still closed the good trade at 80 pips, we would be up $16 instead of -$1

So, I think instead of locking in profit in pips by closing half lots, might be better to just try and lock them in by moving SL up in hopes to close the trade at possibly less pips but full dollar amount at close and gain Cash rather than Pips.

Just my 2 cents
Author:  fxozgirl [ Sat Nov 23, 2013 9:28 am ]
Post subject:  Re: Bob and Shelley Again
matrixebiz wrote:
SteveHopwood wrote:
fxozgirl wrote: This week both my accounts in positive pips but cash loss. What is causing this is trades that I allow to go to full stop loss.
True, we don't want trades to get to full stop loss if we can detect a bad entry early enough and get out of it before it goes too bad but those using the Close Half Lots function may also be cheating themselves out of Cash. I realized this when looking at the History indicator and I was in positive pips but negative cash and I was like 'what' how can that be, until I diagnosed it a little. I'll give a little example below for those that don't understand what I'm talking about.

Lets say I'm trading 20 cents a pip;

Scenario 1 (w/Half lot close) - I've just had a SL of -45 pips then another trade closed half lots at +30 pips and the other half closed at an additional +20 pips. So what just happened is the bad trade I lost 45 pips but positive 80 (30+50) pips in the good trade I closed Half to lock in profit, I'm ahead positive 35 pips, Oh yeah! .... but in reality when you look at the cash, -45 @ 0.2 = $9 and +80 @ 0.1 = $8, so I've actually lost $1

Scenario 2 (No half lot close) - I've just had a SL of -45 pips and if I can get a full trade to close at least +50 pips and not worrying about a +80 like above, I've only gained +5 pips profit but in cash I'm actually up $1 -45 @ 0.2 = $9 and +50 @ 0.2 = $10
So also, if we were in the exact same scenario this time as scenario 1 and still closed the good trade at 80 pips, we would be up $16 instead of -$1

So, I think instead of locking in profit in pips by closing half lots, might be better to just try and lock them in by moving SL up in hopes to close the trade at possibly less pips but full dollar amount at close and gain Cash rather than Pips.

Just my 2 cents
The post you have quoted above is from Steve - not me ;) ...just to be clear for other people reading this thread, my live account gained 2.5% in this weeks trading.

In regards to your comments about the half lots closed, I take your point...when I first started trading BASA I wasn't using Half Lots Closed and had an awful lot of trades that reached 30 pips only to retrace and close at BE+2. Admittedly this was during July/August when the markets are typically slow which I'm sure had a big impact. For me personally, I decided to incorporate Half Lots Closed in my trading and lock in some reasonable profit which I am very happy with at this stage.

This is one part of the settings that is completely subjective to each person's trading style and probably dependant on broker feeds etc which has been discussed in the posts above.

Unfortunately there is no 1 settings that will suit everyone and every broker...so please I encourage you to test this and share your results.
Author:  garyfritz [ Sun Nov 24, 2013 2:56 am ]
Post subject:  Re: Bob and Shelley Again
SteveHopwood wrote:So, we gave the same criminal + 2 live accounts + same bot trading same lot sizes on the same charts = different results. Wow.
Yes wow. I cannot imagine why Empty4 can't manage to produce consistent results with identical conditions. Makes you seriously question if it's just randomly making stuff up, or what.

I come from a Tradestation background, specifically with futures. Futures are a different game, since they're traded on a central exchange. For the most part everybody gets the same data feed; certainly everybody running TS gets the same feed. And system (EA) trading results are 100% replicatable on anybody's TS.

Forex is less consistent because everybody's connecting up to a different broker, and every broker has a different feed. Every broker also has the opportunity to game their feed to hunt stops &etc. So I can understand some differences from broker to broker. But I can't understand why Empty4 can't even be consistent in two identical situations. It's scary, frankly.
Author:  fxozgirl [ Sun Nov 24, 2013 4:52 am ]
Post subject:  Re: Bob and Shelley Again
SteveHopwood wrote:
fxozgirl wrote:
SteveHopwood wrote:Cowboy Cowboy Cowboy Cowboy IBFX last week, all H1:

That was a move of 111 pips in 4 hours. I popped an ATR 14 onto the chart - the reading at the time was 24. Looks like we need a filter to avoid this scenario. I suggest ATR; no trade if the previous candle moved by ATR x an_input_multiplier.

:D
I have already been looking at this too...if you leave it with me, I'm 3/4 the way through analysing all live trades BASA has taken since end of August to see if an ATR filter on the Shift 1 candle would have prevented the big losses.
The other thing I'm checking is to make sure it doesn't prevent too many wins as well.

I should be finished in a day or so and I will post my analysis.
I have already coded a version with the filter. I will hang on until the results of your research to see if it is worth publishing.

:D

I've spent some time analysing all the trades BASA has taken on my LIVE Cowboy Go Markets account to try and identify if there is any value in adding an ATR entry filter to BASA. This has been trading on the 4H chart and pretty much set and forget.

First some actual results...
BASA took 65 trades from 19/8/13 - 23/11/13
51 trades were winners, 16 trades were losers = 78.46% win rate
Total profit = $932.33
I have to say I'm pretty happy with that, but as we have all seen if we have a bad trade that goes all the way to Stop Loss it does hurt.

I decided to see if adding an ATR 14 filter to the entry criteria would help save some of those full stop losses, so using the following criteria I came up with some interesting results.

Criteria: The number of pips in the previous candle (based on the open/close) has to be less than the ATR level chosen for a trade to be taken (assuming all other BASA criteria is met as per EA settings) I then applied 4 different levels to see how this affected trades taken and actual results.

Test 1: ATR x 1
34 trades taken
27 winners, 7 losers = 79.41% win rate
Total profit = $947.66

Test 2: ATR x 1.25
43 trades taken
34 winners, 9 losers = 79.07% win rate
Total profit = $944.54

Test 3: ATR x 1.5
49 trades taken
39 winners, 10 losers = 79.59% win rate
Total profit = $996.09

Test 4: ATR x 2
62 trades taken
49 winners, 13 losers = 79.03% win rate
Total profit = $1128.20

Conclusion:
Using the ATR filter reduced some losses, but there were losses that were going to occur pretty much anyway.
  • * Test 1 prevented almost half the losers but also wouldn't have taken quite a few winning trades, with the end result (in terms of $) virtually the same.
    * Test 2 had the worst result with $ ending up less than the actual account
    * Test 3 had 6 less losers than the actual account, and increased $ a little more
    * Test 4 had the best results preventing 3 trades which were big losers on the actual account and resulting in almost $200 extra profit.
Moving forward, I think the benefits of having the ATR entry filter are worth persuing. Obviously results are going to vary depending on your broker, settings etc, so please I would encourage everyone to do their own analysis.

Steve...if you are willing to incorporate an ATR filter, could you please make it an input that traders can either switch on/off if they want to use it and have a user defined input for the ATR level they wish to use?

Also, I'm sure (hope ;) ) Gary will jump in here with his mathematical genious and share some more insights he may have gleaned from these results :D
Author:  phil_trade [ Sun Nov 24, 2013 8:03 am ]
Post subject:  Re: Bob and Shelley Again
garyfritz wrote:
SteveHopwood wrote:So, we gave the same criminal + 2 live accounts + same bot trading same lot sizes on the same charts = different results. Wow.
Yes wow. I cannot imagine why Empty4 can't manage to produce consistent results with identical conditions. Makes you seriously question if it's just randomly making stuff up, or what.

I come from a Tradestation background, specifically with futures. Futures are a different game, since they're traded on a central exchange. For the most part everybody gets the same data feed; certainly everybody running TS gets the same feed. And system (EA) trading results are 100% replicatable on anybody's TS.

Forex is less consistent because everybody's connecting up to a different broker, and every broker has a different feed. Every broker also has the opportunity to game their feed to hunt stops &etc. So I can understand some differences from broker to broker. But I can't understand why Empty4 can't even be consistent in two identical situations. It's scary, frankly.
Hi Gary, Steve

for information, I'm using this code in my windev/API/Empty4 software, which manage 14 LIVE customers Empty4 account on the same PC, with different brokers, to check availability of broker and symbol feed :

Code: Select all

PROCEDURE BrokerEtDeviseOK( Empty4 , NomDevise)

SI Empty4.IsConnected() ET Empty4.MarketInfo( NomDevise,InfoMode.TRADEALLOWED) = Vrai ALORS
		RENVOYER Vrai
SINON 
		RENVOYER Faux
	FIN
FIN
I use it before checking entry trade conditions and before managing existing trades.

Many time by day (0 to 5), I can receive message from different broker ( also GP ) and different symbol, on my journal tab, saying this broker or this symbol was not available. One Second/Minute after, all is OK.

This should also explain why we can have these various results.

cheers

Philippe
Author:  fxozgirl [ Sun Nov 24, 2013 8:26 am ]
Post subject:  Re: Bob and Shelley Again
phil_trade wrote:
I use it before checking entry trade conditions and before managing existing trades.

Many time by day (0 to 5), I can receive message from different broker ( also GP ) and different symbol, on my journal tab, saying this broker or this symbol was not available. One Second/Minute after, all is OK.

This should also explain why we can have these various results.

cheers

Philippe

Wow, Philippe... pretty amazing that you have discrepancies like that so often! No wonder results are so different
Author:  zTB [ Sun Nov 24, 2013 8:41 am ]
Post subject:  Re: Bob and Shelley Again
garyfritz wrote:
SteveHopwood wrote:So, we gave the same criminal + 2 live accounts + same bot trading same lot sizes on the same charts = different results. Wow.
Yes wow. I cannot imagine why Empty4 can't manage to produce consistent results with identical conditions. Makes you seriously question if it's just randomly making stuff up, or what.

I come from a Tradestation background, specifically with futures. Futures are a different game, since they're traded on a central exchange. For the most part everybody gets the same data feed; certainly everybody running TS gets the same feed. And system (EA) trading results are 100% replicatable on anybody's TS.

Forex is less consistent because everybody's connecting up to a different broker, and every broker has a different feed. Every broker also has the opportunity to game their feed to hunt stops &etc. So I can understand some differences from broker to broker. But I can't understand why Empty4 can't even be consistent in two identical situations. It's scary, frankly.
Please consider the following:

• same broker/criminal;
• same EA;
• same signals;
• same trades; but
• two different accounts;

are quite likely to produce significantly different results simply because of different latency leading to different spread and different slippage. Of course, there may be other reasons.

I am attaching an Excel file which displays the spreads as they change tick by tick to enable one to watch the huge percentage changes that take place in the spreads of lower liquidity pairs even within a few ticks.

To calculate how big is the impact of cost of the trade (spread+slippage), please visit:
http://fxtrade.oanda.com/analysis/sprea ... alculator/

Z
All times are UTC Page 78 of 120

Powered by phpBB® Forum Software © phpBB Limited