Just a quick thanks to you guys for your public blog! I am fairly new at Forex trading, although have traded stocks for a few years. I am trying to absorb as much of your conversations/advise/etc as I can and read it on a regular basis. I has been extremely helpful. Considering downloading your alarm signal also if that is ok.
Thanks again!
Cheers.
[B]TALOND[/B]
Edited last post to, on or about the figures stated. It was on the fly and not pre - set orders. 
Thanks TalonD
You always ask my questions before I can, that saves me a lot of typing.
I’m having a lot of problems with my fingers right now so Thanks again.
John
Hi TALOND
Thought I’d stick part of your message on here to get a couple of ideas going re testing this out with a strategy tester:
When price line crosses the 21 SMA initiate a buy/ sell… no exceptions and no SL or TP. When PA is trending its big pips. The problem is when PA is flat. Thats where I have been struggling. Too fast an MA will cause price to cross further away than it otherwise would on the slower 21 SMA. If price line crosses the 21 SMA and stalls and goes back through the 21 SMA because its so slow its in the same place so all its cost you is the spread. When PA is flat it might do this all day but only cost around 18 pips. But if theres movement you will make small to very large gains. But it needs to be price line and not a fast ma. Otherwise by the time even a fast ma crosses price will have already crossed and if a retrace this will cost alot more pips with the lag.
That is the biggest problem I think…when PA ranges up and down or skirts along the middle band before deciding where to go. One other option I could think of is when price first crosses the middle band and goes up and then retraces to the middle band and bounces off it then we trade in that direction.
The problem with EAs is they take some very direct rules. I think it would have to have a stop loss just simply for the times where price moves past the middle band but then goes backwards.
So, some immediate rules for thought:
1/ BUY when price moves from below the band upwards
2/ SELL when the price moves from above the band downwards
3/ Don’t make any more trades until that trade has been closed
4/ Set an initial stop loss of say 50% of the hourly ATR 25ish pips
5/ Move the trade to breakeven after 20-25 pips
6/ Trail the trade using the middle bollinger band
Alternative SLs are the opposite sides of the bollinger bands.
We’d need to think of a few rules surrounding started trades, ones that failed, and ones that succeeded. Would we wait a few hours until initiating another trade if it failed? What if it had moved down 300pips and hit the middle bollinger exiting the trade. Would we immediately place a buy trade? If not, the EA is just going to mindlessly buy and sell and that is where the losses will come from.
In fact, stick up an hourly chart - everytime the PA crosses the bollinger, look for a time it immediately comes back, would you have bought got stopped out, immediately sold got stopped out, immediately bought again? That’s the sort of stuff you need to stop an EA from doing before running it through a strategy tester.
[B]TALOND[/B]
Obviously I could not know how far PA would go with certainty so by spacing the trades I spaced the risk. Had I entered all four positions at once on my first entry I would be out by around 320 pips! By spacing the trades my max draw down at PA peak was 130 pips’ish. Plus a much increased chance of BE or a bigger take.
Hope that clears things up. Just another way of trading. I won’t have the Vodka but might have a ‘ties’ worth of beer on the way home. 
Don’t think I could hack trading like this but I’m interested
When you said 50% of the move did you mean when you thought 50% of the news spike had occurred? Presumably you’re saying that what goes up in a news spike (in this case) has to come down at some point but you have to hope for at least a 50% retrace to be profitable don’t you?
[B]TALOND[/B]
Well first and foremost the trade has to be contra to the longer term PA direction.
When I placed the first trade PA had already moved up around 50% of the total move. The longer you can leave placing a trade the better. Suggested entries are at S/R levels.
The next 50% of the move was covered by four equal distant trades. But lets keep it simple. Say it was 1 trade at 50% of the move and 1 at 100% of the move. Now PA only needs to retrace 25% of the total move for you to be at BE. Any further down and you make serious pips.
Alternatively, I could have closed the four equal distant trades as PA passed each by the 20 pips on the down. Less combined pips but locked in profit as PA retraces.
[B]SanMiguel[/B]
As they are my words and I’ve spent many hours going over every conceivable combination of ma etc and ultimately deciding a price line was the best option, I’ll add nothing further. :p:p:p
Yes, we were talking about the options you 2 had come up with by PM.
It’s not the actual PA crossing I have the issue with it’s what to do about stop losses at the point of entry.
An EA is so robotic that if price crossed the middle band 10 times in half an hour, it would enter 10 trades. I’m not sure that’s how we’d trade it
I can’t run a strategy test without 100% definite rules. If you want me to test it without a stop loss or anything else then no problem but I’ll put a ties worth of beer on it that it’s gonna come up negative…unless we can define how we would exit a trade in bad circumstances.
[B]SanMiguel[/B]
Bugger aint it… the devils in the detail. I could almost see the ‘Holy Grail’ of EA’s. :D:D:D
Well I’ll go back to the drawing board and see if theres a way around or at least limiting the downside. Otherwise a very good strat! 
I don’t want to pull the thread too far off topic with the EA idea, may want to start a new thread on it? or keep it here? R. hope you’re not upset that I mentioned it to SanMiguel but I get the impression he is pretty good with the mql programming.
there are lots of details and tweaks I’m sure. but I was thinking even if price bounced around the center and you got lots of trades opening and closing, if youre entries and exits are close enough then you might not loose a whole lot more than just the spread and the big moves would make up for that plus some?
makes sense, kind of reminds me of dollar cost averaging in the stock market.
now explain what’s a tie of beer? is that like a six pack? or is that like ‘tying one on’ ?
You already both spoke about it the other week didn’t you? We had a few posts on it I think it’s alright to go through a couple details here as it is essentially MA/bollinger trading anyway - unless it gets too far off topic.
[B]TALOND[/B]
SanMiguel is also perhaps one the best back testers on this whole forum… I kid you not! 
Haha! I ties worth of beer goes back away on another thread. It was a Sunday and we were off topic. We were discussing the finer points or otherwise of beers… and yes Sanmiguel did come up.
Arguably one of the finest beers made in the UK is by ‘Fullers’ a small brewery just down the road in London. They will give you a free tie if you can drink six pints of Extra Special Brew (ESB). I have… errr … more than one or two! :D:D
I need a new tie…
How about packing one up securely and shipping it to me? I’ve tried my hand at making some home brew with mixed reslults.
"Price direction: direction of price within the band identifies convergence-divergence with the current trend, eg price moving up but in lower segment of the band signifies divergence, price moving down in the lower segment of the BBS signifies convergence."
Can anyone please elucidate on this ? i seem not to get what Alan S. Farley was saying here.
Thanks for your help.
John
[B]JOHN4PIPS[/B]
There seems to be an important piece of information missing from this statement. In order to make sense, it should also clarify [B]In what direction the trend is moving[/B]?
Hi Carter,
Thanks for your response.I got this quote from the start of this thread.Please see the whole excerpt.
Excerpt from:
Alan S Farley (2001) - The Master Swing Trader - McGraw-Hill
(note this book is more aimed at swing trading the stocks but some methods apply equally well to forex).
Swing traders must investigate central tendency in every promising opportunity. Market action should spring back toward a centre of gravity after extending in either direction. This axis tends to support price from below and resist it from above during active markets. In flat rangebound periods, price action commonly oscillates back and forth across the pivot until volatility triggers a new directional impulse.
BBs focus analysis of central tendency in real time. To use the bands effectively, apply a central moving average that tunes into the expected holding period…
Also see: Bollinger Band Tactics by Alan Farley
Simple band concepts:
[B]- Price direction: direction of price within the band identifies convergence-divergence with the current trend, eg price moving up but in lower segment of the band signifies divergence, price moving down in the lower segment of the BBS signifies convergence.[/B]-
Trend testing: the lower, centre, and upper bands represent S/R for the trend. Reversal off any band increases odds that price will expand in the reversed direction and return to the last band crossed or touched.
2. Penetration through the centre band increases directional momentum:
Trade management:
Stop losses on the counter trend trade - move to breakeven after 25 pips particularly on the counter trend trading but this is up to you personally.
Stop losses on the trend trading - opposing bollinger or candle wick.
Target is the middle bollinger on counter trend trading.
Target is until a significant S/R zone or until the MAs cross back over in the opposite direction.
Probably no need to trade the trend and a counter trend trade, just 1 or the other depending on the setup.
Quote:
Originally Posted by R Carter
When entering any trade I look to the previous high or low (usually a wick) and place my stop just outside of this. I like to manually move my stop inside the trade as soon as is practical. This largely depends on how fast price is moving and what size the retraces are. I then move the stop up/ down the trade as it progresses in line with the above movement.
Trading times:
Potentially any high liquidity sessions but prefer London - NYSE session
Inherent problems with MA indicators:
Whipsaws - the MAs are primarily used for trend trading on the 1hr chart. If price is ranging you are going to get whipsawed in and out of trades. This is why we use the bollingers to give us an indication of probable price. If price is ranging, then we can use the counter trend trading instead and look for “Head in Ceiling” or “Foot in Floor” trades.
Lag on longer term MAs - all MAs have a built in lag but these settings should be good enough for intraday trading.
Whipsaws on the counter trend trade - entering a counter trend trade from the bollinger edge and wick candle only to be caught in a continuation of the trend.
The standard setting for bollingers is 2 std deviations - if you want you can wrap a few bollinger bands around it or fine tune to your chosen pair but should see price rarely hit the 3 std deviation bollinger.
Ultimate price movement will depend on all S/R boundaries and not just those associated with BBs.
It simply means that in a strong uptrend, you will mostly see all the candles and price action in the top half of the bollinger.
In a strong downtrend, you will see most of the candles in the lower half of the bollinger.
For example:
1/ if you are in a downtrend but candles have crossed over from the lower half of the bollinger into the upper half of the bollinger, price is diverging from the trend and slowing down;
2/ Candles are moving sideways or upwards but are currently in the lower half of the bollinger - similar to above but hasn’t crossed over middle band yet.
Bring up a 1hr chart and find and obvious up or down trend, which half of the bollinger is all the price action in?
[B]SanMiguel[/B]
Thanks for the chart it makes things crystal clear… for price to be diverging when moving to the centerline of the bollinger the trend had to downward but this was absent from the explanation alone.