Saturday, 21 November 2009

Stop loss placement - a specific solution.

One of the biggest problems facing a trader is what to do if a trade goes wrong.

Do you trust your original judgement and hope it turns?

Does your fear of being wrong cause you to close it too soon before it does significant damage?

With WMs, last December 2008 on the DOW, was a nightmare. They kept forming on RSI, then Price, then retracing and taking out the base and then reversing again, lovely trade IF ONLY I'd had stayed in a bit longer.

Problem is, how much longer is "reasonable"?

This idea is based upon a specific situation in EWT as posted by SED on the iii UKX board - this is my interpretation & I have added in the WM idea.

Wait for a clear wave1 with 5 wave substructure.

Wait for the wave2, with a clear 3 wave substructure,to retrace at least 38% of the wave1.

Wait for the WM to form on RSI and establish the base level on Price.

Split the trade into chunks so that the S/Ls take into account the base level on Price and at least two of the Fib levels below that for a W.



I realise that the 80% is not a true Fib level but it is the figure use in a lot of EWT research as signifying a boundary between expected and unexpected. As usual the conditions under which the data was collected is very important, as is an understanding of what the researcher means by "unexpected".

With this example I'd split into three chunks and use below the base, below 62% and below 80%.

It increases the cost compared to just using below the base BUT it adds a reasonable amount of extra space. [IMO, never commit more than 2.5% of your bank to any one trade.]

cheers theory

WMs on RSI in context of EWT waves.

I think of three types of WM "failures", where probably only one of them is really a failure, one is disappointing and one can prove very useful.

The real failure is where it forms on RSI, maybe even Price and then collapses. The only defence being phase out and exit.

The next one does form on RSI, Price and even moves far enough to set up a small loser, break even or maybe small profit before it reverses.

The final category forms on RSI, Price and delivers a good return before retracing and taking out the original shape on Price.

I now feel I can go some way to placing them relative to EWT waves. I am not saying that all EWT waves end on a WM, just that some do.

The purple waves with the ? after them are the unknown next phase.



The failures are more likely to be in the following positions:

at the end of the 1st wave, if the 2nd turns out to be mild;
at the end of the 3rd wave, the 4ths are usually mild;
at the end of the 4th wave, if the 5th truncates.

The disappointing ones probably are:

at the end of the 1st wave, if the 2nd turns out to be neither mild nor sharp;
at the end of the 3rd, if the 4th puts up a bit of a fight but then gives up.

The last category seems to me to be best located:

at the end of the 1st, the 2nd is a high % retrace and then wave 3 kicks in.
at the end of the 3rd, the 4th is a high % retrace and then an extended 5th occurs.

It is possible to get a good return from trading a smallish % retrace of a very big wave OR a big % retrace of a smaller one. However hitting a big % retrace of a big wave is a much easier way of making money.

I hope this helps you with "shot selection". Previous posts have looked at the context of WMs in terms of RSI level, Bollinger, Trendlines, R/S levels and now I am adding EWT waves.

cheers theory

Saturday, 31 October 2009

Work In Progress - ongoing development.......

Rather than replicate everything in two places, I have included a link below to a newly formed discussion site I have joined.

The III UKX site was no longer providing what I needed - end of story!

Aplogies but cannot get the link inserter to work, so copy and paste is the only way around it.

http://ftfexcel.invisionzone.com

The work is in the FTSE folder under the name of "theoryman". Guests have the full access to images.

cheers theory

Sunday, 13 September 2009

Linking ideas together - WMs, EWT & Channels

Just to let you all know the direction I am moving in.

I'll be trying to use EWT from a short time frame basis, checking possible patterns & then putting them together to make vaild possible counts on th longer timeframes.

I'll be using two ZigZag indicators, one set on High & the other on Low to mathematically identify the key Highs & Lows needed to decide if it's moving in 3s or 5s etc.....

I don't think I will bother to fully automate the process but I'll be using some code to try and remove the "seeing what you want to see" problem with any pattern spotting.

cheers theory

Saturday, 15 August 2009

More questions than answers [EWT again!]

As I posted on the iii UKX board, I took yesterday off, so this is a reflective/hindsight view. It has however helped me with my attempts to link WMs on RSI to EWT on Price. [Anything to do with EWT thoughts as opposed to the system is in square brackets.]

Firstly I'll compare and contrast the two Ms in terms of the system. One is above RSI 70 and is formed when the Bollinger bands are wide apart. The other is just above 60 and is formed when the Bollinger bands are close together. They are both "fallers" - the first one only just.They both are OK in terms of Price action /Bollinger. The first is tradeable, the second is NOT!!! Why not? Answer comes later on.





[I feel that both clearly come at the ends of EWT waves. The first one at the end of a trender, the second at the end of a corrective. I already "feel" that correctives can be gentle or sharp, this was a gentle one. The Bollinger bands signal what has gone on before, the Ms the end of waves.]

[Now some real problems for my system!]

Because the corrective went above the B/E for the first M, I'd have been stopped out on my backstop. The only way I could have survived is to keep some back at the original S/L.

[My best WMs, in terms of meeting the rules, come at the end of trends, so am I just trading the correctives? IF it's a sharp one then bingo but if it's a gentle one I only get my "0.5(20+0)" sort of return.Should that concern me, still making money?]

[Some of my best WMs in terms of return have been the taking out of a previously verified R/S level, the infamous "failed WMs etc....". So that's a real bonus, the system indentifies a corrective, which then fails, so it leads to a motive.....]


[The second M was not a system trade, fails the Price/Bollinger reward:risk check. If it's a gentle corrective then the Bollinger bands must be narrow but that doesn't mean the next motive won't be a real trender.]

So in a nutshell, the reward:risk rule is working both for and against me. It filters out those WMs generated by market noise BUT it also keeps me out of those at the end of gentle correctives which MIGHT lead to a really big move.

I need time to think + would welcome any thoughts from practised EWTers.

The next image shows a wW which is uncommon.

[Signalled the end of the powerful wave 3? The subsequent move back up to the region of a standard Fib level(not on image) seems to confirm that.]



---------------------------------------------------------------------------------
Finally, for those who like 1-2-3 trades on Price, this was the set up that coincided with my 60 min M that was NOT a system trade - maybe that is something that needs looking at?



cheers theory

Thursday, 13 August 2009

Should have done better and some [EWT thoughts].

The first image shows a sequence of events on the 14 min. timeframe.

The first one is a straightforward trade. The W then kept appearing on longer and longer timeframes, which I usually find is bullish - OK it was a slow burner but it finally delivered.



Here is what the 40 min looked like when it formed.



Now back to the 14 min., same chart as the top one, just placed here for ease of access.



It delivered more than it looks like it did because the post 16:30 candles were reduced by around 8pts on the IG cash chart, to allow for the dividend correction.

So its base is a verified support level. When it failed next day, I went Short, expecting a new EWT wave down. [ The W IMO signified the boundary between two waves, the lack of oomph after it formed led me to believe it was a corrective, so we should get a nice motive down to follow.]

What I didn't take seriously, I just wasn't in tune with the market that morning, was the second W which formed on the chart. [Now seems obvious that the wave down to the first W was a 3, followed by a corrective 4 and then a weak 5, the end of which was signified by the second W.]

The internal structure of the second W is better than the first on RSI. Just check how chunky the bodies of the candles were on Price in the second compared to the first one.

[Note to self, wave 3s can never be the shortest in an impulse wave but wave 5s can be!]

RSJ on the iii site called it perfectly. I only joined in later on the breakout OCO after the continuation W had formed.

That trade delivered the famous 0.5(20+0) and then I got a trend continuation that behaved itself perfectly.

It led to a possible EWT count that really fascinated me. The possibility became more and more probable and it was a very enjoyable experience. Again, thanks to RSJ for confirming that my count was a possibility and not some rubbish from an EWT novice.

I am convinced that the next big step forward with this system will be melding it with EWT.

cheers theory

Saturday, 1 August 2009

New idea + old faithfuls

OK let's start with a standard trade.
If you need the practice go through the checklist:
RSI shape,
RSI level,
Price/Bollinger,
Possible initial reward:risk assessment.

Apologies for the next bit two bits being rather "wordy" but I can't think of a way to avoid it.

I then have added a new idea. It's a standard support level as shown on the diagram as the green line BUT it's validated by the trade following from the W delivering the 0.5% return or better. This validated level is then the basis later on for a bounce OR crash through trade. I have been focussing on the unexpected crash scenario (more fun) and kept missing out on the expected bounce.

It was the market behaviour at 4500 which made me realise I was missing out. The top of that 1Hr M held out many times, OK I got the crash through spot on but opportunities were missed.



On the next diagram the support level appears as double thick green dashed line, the double bit will become obvious when you look at what happened.

The black lines are downstream channels based on previous Highs. The lower ones are the standard parallel click and drag through previous Lows.

Apologies for the red lines they are part of something I am trying out.

So I was ready for either the bounce or crash scenario. Once it was a confirmed move below the support line AND break through the trendline, I would go Short going with the double whammy. If, however it bounced and went back inside the channel then I'd go Long. The support was confirmed so the line was made double thick.

The thin green line on the next chart shows the PDH used during trading the following morning. During the night the index moved way above where I was expecting it to be - nice surprise. Once it came below that level I put covering trades on and so was neutral. There was no specific evidence to close the trade but I was looking at a healthy profit. The light blue trendlines showed the possibility of more to come BUT the early moves to close gaps and high volatility make this my least favourite time to be running a trade. The purples ones were added later.



So what made me remove the cover? Well it was the weak signal from the W in the next image. No way is this a system ENTRY but it was good enough to warrant letting the trade CONTINUE. If need be, go back and re-read the section on trend reversal shapes and trend continuation ones.



Trigger on the UKX board and of "moo river" fame, would be proud of what's in the next image. The upstream crossing of a downstream river and a breakout.



Finally just a few simple trendlines, a continuation W where the RSI signal that confirmed it had formed, was as high as any we have had. It delivered 20+ on Price, with the RSI hitting a maximum afterwards of nearly 80.

Key point here is that for continuation WMs it's the shape and NOT the level that matters, plus the Price/Bollinger is totally ignored - unless of course you include the Price moving along the SD=2 band!



cheers theory

Wednesday, 15 July 2009

Third trial Cable trade.



The W has 7 points, as referred to in the first post in this blog i.e. Stage 1.

My software identifies the Pt 1, 2 and 3 - as used in the pdf on 1-2-3 trading. Point 0 is the one before Point 1 and Point 4 is the trigger one on RSI.
The least number of dots is therfore 5, Points 0,1,2,3 &4.

This shape has Points 0,1,2,2a,2b,3 & 4 i.e. 7 in total. What are 2a and 2b? They fill in the gaps between Points 2 & 3.

I prefer my WM shapes on RSI to have as few points as possible and 7 is my maximum.

A quick run through the checklist should confirm the trade meets all the rules.

The following images show how I use trendlines. I prefer to think right to left, so my trendlines are based on the latest information. In the case of an uptrend like this, the latest two Lows form their basis.

The single black line was all I used to start with. The initial two Lows define its slope, the third touch confirms it, the breach calls for a re-think.

The horizontal lines in red are my S/Ls, my style is to split trades. The upper dotted line is the "snuggle up" position and the solid line the "time to close down" level. I want to allow the trade as much freedom as possible but without throwing profit away, on allowing too big a retracement before closing it down. They are positioned below both a previous Low AND the trendline.



The latest Low is then used to construct the next channel. This time the upper trendline, formed by click and parallel drag on the chart, fits data that is already there, so it now goes in double thickness and I chose red to show up well in the posts.



After that I added the next trendline, using different colour and design of line. The problem becomes having too much info but this is a trial trade, so what the heck.



I have included RSI14 in the image below, using the standard line setting. This shows a W on RSI, of the continuation type, forming when the Price bounces off the already established red trendline again - context,context, context!



The split trades finlly finished when 16300 was hit and then finally 16275.

cheers theory

Friday, 10 July 2009

Cable trades - widening the net.



The first M, in red, was a valid trade. It was possible to get to B/E and wait for it to deliver. IMO it didn't deliver enough to warrant being rated as a potential "failed M etc...." trade. On the FTSE I use 0.5%, which on Cable equates to around 80pts. Trading the breakout through its top would have been over aggressive IMO based on FTSE findings.

The second M is a cracker. I've included the S/L, entry and initial target on Price. The reward:risk ratio is clearly OK. Make sure you go through the checklist for all the other rules.

When there was a chance of the W forming I moved into defensive mode and locked in profit. My exit would have been on the W forming on Price AND moving within the SD=1 Band.

Ws that form on the SAME time setting as the original M are rated as big dangers. I forget the rules for entering a trade, if they form on Price, I get out whatever their quality.

IMO there is a big difference between putting money from the bank into a trade compared to being too casual about conserving profit from a runner.

The trade survived but later another W started to form on 25 min., and I took the view that the market was having another go at reversing. I waited for it to form on RSI and go back above the SD=1 band, got stopped out. IF I had sat back and waited for it to form on Price, like I did with the first one, then I would have survived.

Now for the second trade, in which I did not cover myself with glory.



Inverted H&S, only the black W is traded, no anticipation!

Entry 16190, S/L 16150,initial target 16250. Reward:risk ~ 60:40, OK.

Quickly go through the other rules, no problems.

I just didn't like the way it panned out. Seemed to keep hitting its head on around 16210 and it broke down through a wedge I had drawn on the 5 min chart. So I closed it for a loss of 13 pts. As soon as I did that it started to go back up again, got to B/E and into profit......

I am still very much a novice at trading the Cable since this was my first one! It is a long time since I got caught out like that on the FTSE. I might have closed down half on the FTSE but not all of it, the base of the W never came under threat.

On reflection it was very poor technique, I got drawn into action by the 5 min Price action and forgot about the biggger picture i.e. the 45 min one.

cheers theory

Thursday, 9 July 2009

Why isn't this a tradeable W?



Run through the checklist:

RSI Shape is sharp and tight, good internal gaps.
RSI min. is only ~40
Price/Bollinger is great, a double down and up again
Entry is at RSI around 47 or 48 giving a Price of 4164 to 4166.
The initial reward to the middle Bollinger is less than 10.
The known risk to just below the base is more than 15.
The reward:risk ratio is less than 1.0, never mind 1.5!

The bodies of the candles are chunky and alternately coloured. However, the volatility within each candle, means the tail on the one that matters, in deciding the S/L, makes the risk too high to warrant a trade in this case.

IF the middle Bollinger line had been say 25 away, then go for it.

cheers theory

A lot of effort .......

Last night the readings of RSI were below 30 on many timeframes after that fall. I was scanning the standard settings 15min, 30 min, then I tried 31 min, just to have a look.



Notice the internal candles are Red Red Green Green. Nothing wrong with that pattern at all BUT I do like it tighter IF possible.

Now if you double the time setting then either the first two pair up and then second two pair up and give really big internal Red Green candles combination or the inner two do and cancel each other out - doji formation time.

Quick look at 62 min. and it's obvious what has happened this time. Beautiful chunky equal sized alternate coloured candles, no wonder the RSI W is so well defined.



Straight forward trade, check the filters for yourself, it should be coming automatic by now!

Sharp tight shape?
RSI level?
Price/Bollinger?
Entry RSI will be? Leading to price entry @ ?
Reward:Risk?

So today it was fixed in my mind that the signal was on ~ 60 min, so when the M came on the 14 min, it was just "noise" and I defended against it without problems. When the W on 14 min arrived I was in trend continuation mode.




The W on 18 min reinforced the trend continuation in my mind, the later M was a good quality one and just in case I shifted the S/L on my running trades to where it would form on Price. It did and with a vengeance.




Why the title? Well it was all building up towards a taking out of 4200, which was just above the top of the M on 14 min that delivered big time yesterday. That would have been a previous strong M, that really delivered but then failed, time to launch some Longs from the profit of the ones already running.

An awful lot of planning and thinking out possible alternatives went into the day's trading. Just seemed at the time like one very big let down, by tonight I'll be thinking of the profit banked!

cheers theory

Wednesday, 8 July 2009

This one didn't get away!



The first W from last night met all the criteria except one, the reward:risk was too low, tempting but dangerous.

The next one was a horrible shape internally - no shifting of time setting seemed to give a sharp W on RSI.

The H&S on 14 min RSI was an absolute gift, only the black M is used but the symmetry of the overall shape is beautiful! There was also an M on 16 min if you prefer to stick to 15 min or above. Remember most of the system is rules, this part is a guideline.

Max. RSI is above 65.
Single move of the Price above and back below Bollinger Band.
The entry on price was so close to the top of the M, the reward:risk was fine.

For those who follow the iii board, I was already Long from 4162 (non-system trade). I used the shape to close the running trade for just less than 30. I then traded the M from afresh - there was no way I was going to get caught out two days running!

cheers theory

Tuesday, 7 July 2009

Was this a missed trade?



The Price/Bollinger is a double break through and back under.
The RSI shape is tight and sharp.

The max RSI is only above 60 rather than 65 or 70, so it's only in Group C but they can and do work.
It is a faller on RSI but only from 63.2 to 62.9, still counts as a faller though.

The distance from the entry level on Price to the top of the M is greater than the distance to the middle Bollinger line - this is what stopped me. The initial potential reward:known risk ratio was below 1.

How do I rationalise missing out on a trade like that? It's simple, it didn't meet one of the criteria, end of story!

cheers theory

Inverted H&S on RSI



Since I NEVER trade falling Ws on RSI, it's the one in black that is used to trigger the trade.

The lines of best fit in red, on RSI,show the trends before the W that help to make up the inverted H&S.

The W passes all the filters:
tight sharp shape with good gaps between the internal RSI points,
low level of minimum RSI at less than 30,
double move below the Lower Bollinger on Price,
potential reward:known risk ratio exceeds the minimum required.

The horizontal line in red on Price was where I placed my S/L on the running part after the previous move up.

The lower window shows the drop in Price required to produce an RSI of 50, during the active candle, just something I wanted to keep an eye on.

cheers theory

Monday, 6 July 2009

A tale of 4 Ws

The attack on the long running lower channel on the FTSE, using the April lows, was a sign to either trade its cataclysmic collapse or use it as support and look for Ws as the trend reversed.



The first W formed on RSI as shown by the segment but eventually failed as the base on Price didn't hold - can't win them all. A nice high prob. low risk trade BUT it was not to be.

The next one did deliver - standard trade for followers of the system.

Now for the interesting bit!

The continuation W in black. Note the rules for continuation Ws are totally different to those used to find trend reversal ones. Basically forget the RSI level,check the the RSI shape is tight and sharp with good internal gaps and hopefully find a bounce off the MIDDLE Bollinger line. These points have been covered in a previous post.

Since the bounce off the middle line is a key requirement, it acts as the get out level if things do not go to plan. A low prob. low risk trade hoping something unusual happens.

Then something happened that I cannot remember happening before, the continuation W got an extra bit added on its end, shown in red. What I refer to as a Ww shape to reflect the relative sizes but really they are conjoined Ws sharing a common section.

The extra trendline on the Price window was put in to help me locate an alternative closing level should I decide to not use the middle Bollinger line - basically running along the same line anyway.

So now it's just a matter of waiting to see where it goes.

cheers theory

Friday, 12 June 2009

Time for a break

Hopefully most, if not every, possible point has been covered in:

Stages 1 to 6 posted in May,
Worked examples,
Comments.

I'm taking a break from trading, need to recharge the old batteries, hope to start the blog again in early July.

cheers theory

Thursday, 11 June 2009

Mm on 14 min - reward:risk was 1.5:1

The shape was there on the 17 min. but not as good in terms of internal gaps on RSI, so I used the 14 min. one. The windows have been adjusted to the same size for ease of use on this blog.



The long left leg of the M is purely there to define the context I am seeing it in i.e. the possible end of a long running uptrend - it does not affect the Pt1, 2 etc...

Entry was at 58.9 on RSI, Price equiv 4471/0.

The middle Bollinger line was close BUT the risk, to the top of the shape, was very small. Reward:ratio was just above the modified value of 1.5:1.

No problems with the rest of the filters, go for it.

The predicted problem occured, it got through the middle Bollinger and then came back before I could take 20 with half, so I was forced to take 15.

The rest was run from 10 above B/E, which was the top of the shape.

I now had a situation where I was "being paid to trade", the worst outcome was 0.5(15-10) for a 2.5 profit.

The extra little m (last bit of the original blck one + the extra bit in red ) appeared, usually a bearish sign, very encouraging BUT it didn't turn out that way. The trade was still running though when I shut down the platform.

cheers theory

Wednesday, 10 June 2009

Patience rewarded - M on the 23 min.

Apologies for the different fonts, had to add a bit to the original, via Paint, when pacman raised a question via the iii UKX board.



The candles in the M do look much chunkier on Price when viewed on full screen - go and have a look if you don't believe me!

The difference between Pt2 and Pt3 was just over 1, so it just passed the internal check. With all the other WMs so far, this has not been an issue.

Nice high max RSI but only a touch on the Upper Bollinger by Price.
Entry was from RSI hitting 71.1 i.e. 1 below Pt2 which was @ 72.1.
Price equivalent just below 4500.
Risk very small - look where the top of the M is on price.
Reward potential very good - look where the middle Bollinger line is.

The rest is history, IMO a very simple trade to run whatever your personal style.

How about loading up the 23 min chart, it should still be there today and tomorrow.
What happened when it first hit the middle Bolliger line?
What happened when it hit the Lower Bollinger Band? And after that.....?
Put some Fib levels on using the maximum drop from the top and see how they could have been used to run the last bit of the trade.....

With the benefit of hindsight, we can all maximise the return from a trade. In reality it's about getting the balance right between protecting profit and giving it room to breathe. I like to split my trades and give myself the chance to do both.

cheers theory

Anatomy of a losing trade on 16 min



Looked superb, RSI level, Price/Bollinger,low risk, high potential reward.
Entry was at 72.2 i.e. 1 below the level of Pt2, this gave a Price equivalent B/E of around 4474/3.

After it failed in two attempts to get beyond 10 pts, I grabbed 5 with half and left to rest to survive or die, it died.

The double doji in the candles after the one that triggers the trade is not good news at all - this is supposed to be a momentum trade. If in doubt, grab profit with some and keep the potential loss as small as possible, is my personal style.

Total return 0.5*(5-12)= -3.5 pts, I can live with that as a failure!

cheers theory