Trading Systems in 'New Concepts In Technical Trading Systems' by J. Welles Wilder

[Bocajunior]

Originally Posted by sirkeen

IF we should stop trading when the ADX is below DI+/- why is Wilder opening a position on DMI Chart (page 49) between June 20th, and July 13th, even though the ADX is below the DI+/- lines....?

Hello all, Sirkeen.

I was also wondering about the positions you mention. He is also opening these positions when the ADX value is less than 20. Any ideas anyone?

Randont, The geek,
Thanks for posting your trials and tribulations during the last trading week. I find it very interesting to read about all the different emotions that everyone goes through. Keep up the good work.


Best Regards
Boca

[sirkeen]

Originally Posted by Bocajunior

But as Dale mentioned, It is no good to just be rangebound,, we need a little bit of up and down movement (ATR) so we can make a bit of money. So surely the ATR comes in to play here..

Volatility is the second most important factor when calculating the CSI. And therefore, you are absolutely correct in the above statement: The ATR does come into play but IS factored right into the CSI.

IF your rangebound pair/commodity is also giving you a healthy reading (compared to other rangebound pairs/commodities) on Wilder's CSI scale, the potential profitability of the rangebound system you use is increased!

That is the beauty of Wilder's book. Everything you need (with some minor modifications for use in today's markets) is spelled out in 141 pages of pure brilliance! We all can make money using this system, its just a matter of patience, discipline, and whole 'ton' of effort!

M.

[dpaterso]

Good (Sunday) morning all!!!

Regarding ADX/ADXR/CSI:

My 'take' (based on the statement on page 111) is this:

The ADXR is used to 'rate' the DIRECTIONAL MOVEMENT of the different instruments or pairs against each other but DOES NOT take volatility into account.

The CSI is used to 'rate' the DIRECTIONAL MOVEMENT and the VOLATILITY of the different instruments or pairs against each other.

The CSI does indeed solve our 'age old problem' of 'all pips are not created equal' IF set up and used correctly. This is interesting because I've have indeed on occasion set up the CSI for WAY different instruments or pairs to see the effect that it has. For example: you may look at the ATR of a pair like GBP/ZAR and think to yourself that this pair has LOADS of volatility so therefore it must be 'the pair' to be trading. However: if you factor in all the data required by the CSI you'll very quickly see that the CSI 'takes it down a notch' i.e. the CSI may indicate to you that although GBP/ZAR has the highest ATR and may APPEAR ON THE SURFACE to be 'the pair' to be trading it may or may actually BE 'the pair' to be trading because you may find (as I did at the time) that a pair like GBP/JPY is a FAR BETTER pair to be trading EVEN ALTHOUGH it has a far lower ATR value. As you probably know at one stage I experimented with just multiplying the ADXR by the ATR but I have now come to the conclusion that the result of this 'equation' is pretty useless e.g. a pair like GBP/ZAR WILL ALWAYS come out on top and I know from experience that this CANNOT be the case ALL THE TIME. SO: it's either the ADXR that's used OR the CSI 'proper'. The BIG problem with using the CSI is that it's really a HUGE effort (it's not the effort that worries me but the time it would take me every night to adjust the margin requirement for each instrument or pair) because you have to remember at Delta the margin requirement changes every minute based on the exchange rate and the value per pip movement is always different for each pair when converted back to your account's base currency. This is not so much of a problem at a broker like GCI because the minimum margin requirement is always the same i.e. the lot sizes are 'fixed' in $25 increments at 400:1 for instance. For example: RIGHT NOW a single lot of the 'full' or 'proper' Dow will cost me $592.35 at Delta and the value per pip movement will be $1. The margin requirement will change as and when the Dow moves during the day. At GCI on the other hand a single lot of the Dow Futures will cost me $25 and the value per pip movement will also be $1 BUT the margin cost will NEVER vary. Now you could say that this is of little or no consequence when you're dealing with forex pairs because of the 'fractional' price quotes. I don't know: it's up to you but you would then need to work out a 'baseline' or some sort of 'average' for the margin requirements for certain groups of pairs e.g. you'd say that any USD/??? pair has a margin cost of $5 'across the board' and any GBP/??? pair has a margin cost of $9 'across the board' even although this will change during the trading day. Then of course (and this is important) you would also have to work out a 'baseline' or some sort of 'average' for the value per pip movement for certain groups of pairs on the same basis even although, again, this will change during the trading day when converted back to your account's base currency. None of this of course would be a problem trading commodities (except of course for the margin requirement that would change from time to time during the trading day). Also remember that you have to 'adjust' the ATR values for forex pairs e.g. you cannot simply take the ATR being indicated for GBP/ZAR and use that 'as is' along with the ATR being indicated for a pair like GBP/JPY i.e. you have to 'adjust' the ATR so that you're 'comparing apples with apples' otherwise the CSI calculation is going to be useless once again. In other words: you have to multiply the ATR of GBP/ZAR by a value of 10 000 and use THAT value in the CSI equation and you have to multiply the ATR of GBP/JPY by 100 and use THAT value in the CSI equation.

By the way: it is interesting to note that the equation for VOLATILITY is EXACTLY THE SAME EQUATION used to calculate the ATR!!!

As far as 'the man' opening positions when the ADX is below +DI and -DI the ONLY possible answer I can come up with is contained in the fourth last paragraph on page 48 ('... Suppose for instance, as a long-term trader ...') although I'm not sure that what I'm saying is correct i.e. the fact that he is opening the positions in his example on the next page is confusing to say the least. I mean: I would have (possibly) understood if they were all SAR's but on page 49 it is quite clearly indicated that he ENTERED on 28 June and I can not give you a REASONABLE explanation for this based on the text. The only other POSSIBLE explanation is that Wilder, KNOWING that when ADX is below both +DI and -DI it WILL EVENTUALLY turn and rise above +DI and -DI, opened the position but I'd say this wasy risky at best i.e. who knows in which direction the price will go at this point in time. (To me that would be like selling JUST because STOCHASTICS was overbought and we ALL know where THAT one gets us)!!!

NOW:

I need to ask a favor:

Would all the people who have 'the book' and my indicators please be so kind as to send me details of WHERE THEY ARE and more importantly in which time zone they reside (or are trading from or whatever). The reason I ask this is because I'd like to draw up a 'table' of who is where and what their local time is compared to mine so that I can (hopefully) come up with a time that will suit most people (and myself of course) to be online with Yahoo Messenger everyday for discussions, problem solving, support, questions about the indicators, etc. etc. etc. Please email this information to dpaterso@forexbrokersonline.net so I can try to work something out for all of us.

[midulster]

I know I said I would leave it alone but I cant help myself!

Hi my name is Derek and I'm an analyst.

Dale: You got me thinking about the difference between the smoothed and unsmoothed versions of the ADX calculation. I hadn't placed any significance on it when I was replicating his worksheet in Excel but he uses a 2 step smoothing process – first in calculating +DM14 and – DM14 and then those smoothed answers are used in calculating the ADX through a further smoothing equation. So I went back to Excel and removed both sets of smoothing and replaced them with straight 14 period averages – guess what – I get the same answers as you! (Well close enough for jazz anyway)

So I think the smoothing just 'cools' the movement of ADX, not so many extreme highs and lows, it doesn't really alter too much which pairs come top and which come bottom if all you doing is filtering for other systems, so that's why it hasn't affected your trading success.

I think it would however make a difference to trading the DMS and I suspect your new indicator may actually beat 'the man' here as it is probably a little more responsive in catching the moves. It might be worth considering recalibrating the 25 and 20 levels though, to say 40 and 30 to take into account the larger moves. Its seriously crude and I know it doesn't necessarily follow in logic, but he says markets tend to trend about 30% of the time and consolidate about 70% so maybe that could be used as a rule of thumb for setting the equivalent of the 25 level.

To any eagle eyed spotters among you, you'll notice the figures in the last 2 columns for my ADX and ADXR are slightly different to my last effort – it is because I had skipped one day of back data, 18 June. It hasn't made enough of a difference to alter any of the filtering signals.

All this may be useful or it may be the irrelevant, inane drivelling of an addict.

Derek – taking it one day at a time.

PS My ears are ringing this morning after a great night's craic seeing Def Leppard in concert. At least unlike myself they are managing to hang on to their hair!

[dpaterso]

Hi Derek,

NOW you've got my attention!!! Def Leppard!!! (Well it's not my 'band of choice' BUT it's MUSIC and THAT is what keeps me going and has my whole life man)!!! As you may or may not know 'my bands' are: Nightwish (the OLD Nightwish with 'my' Tarja) (although the new 'hits' where Anette is CAPABLE of hitting the notes are fantastic as well as long as I don't have to WATCH her), Lacuna Coil, After Forever, Within Temptation, and Evanescence (and OF COURSE from my YOUNGER days Ronnie James Dio, Black Sabbath, Gary Moore who I KNOW you MUST like i.e. 'just a feeling' given where you're from, Whitesnake, etc. etc. etc. You get the 'genre')!!! To date I've only seen Evanescence live and it changed my life man!!! (I cried like a baby on the first note)!!! If there is ONE PROMISE I HAVE MADE TO MYSELF AND WILL KEEP it is that once I've settled all my debt and reached my financial goal next year I AM going to go to every single live concert I can to see 'my bands' live (as a matter of fact all 'trading things' being equal I intend to put on my OWN concert with my choice of bands over here because with the exception of Evanescence these bands WILL NEVER be invited to play here and they have a HUGE following over here so it amazes me). Anyway: I intend to go to Graspop, Pinkpop, Donnington Park, and any other 'pop' or 'park' you can think of next year to see 'my bands'!!! (What other 'business' would even afford you the opportunity to even DREAM of doing something like that I ask you)!!! (Oh: and I'm forty three years old this coming Monday i.e. tomorrow so this is not some 'teenagers pipe dream' I assure you)!!!

WOW!!! Sorry!!! Off at a tangent!!! I just get all excited when someone talks about music (well music that may be of a 'similar' genre to what I 'live and die for' anyway)!!! Besides: seeing Def Leppard IS a 'big deal' in 'my book' anyway!!!

Back to the land of dreams:

Well I'm happy now that YOU are happy that we have some 'consensus' of opinion on the ADX/ADXR values!!! I DO believe it's going to make a big difference having the CORRECT ADX/ADXR values in front of us. To be honest: I always looked at the DMS and tried to follow some trades through but I always 'came up short' as it were and could not quite see how it could be profitable. BUT NOW: getting the signals ON TIME appears to make a HUGE difference so I'm 'dying to try it' now!!! I DID promise to try to add Wilder's 'smoothing' to the indicator (well for Delta's platform anyway) but I've not gotten around to doing it this weekend BUT I'll get to it (but as you note: I question the difference that it MAY make overall)!!!

[kaalilaatikko]

Originally Posted by midulster

Hi my name is Derek and I'm an analyst.

Ouuch ... that hit me ... ROFL! Should I make the same confession?

I have had a couple of hours this weekend to study ADX and VS, and I am now a bit confused. I'll write down some of my remarks when they are fresh right now, and have a deeper look at them in the coming days.

I've incorporated both smoothed and unsmoothed calculations in my program. It seems that they give somewhat different results, larger than what I expected, but I suppose that they could be taken care of by some sort of recalibrating, as Derek suggests.

Now I have been using Delta's data that spans about 4 years backwards. It seems that the VS parameters that I earlier calculated with 10 years of data may not be the optimal ones when looking at a shorter time. That's something that I have been thinking also earlier, but my big question is here: how much time would be appropriate to look back when backtesting? 10 years may be too much, but 1 year too little.

I experimented by placing ADX checks on the trades that VS selected: trade only when ADX is above 25 and +DI/-DI are in correct order. The results were really confusing. For some of the pairs ADX selected the trades pretty well, but for most of them it seemed that I would do better without ADX. I also tried this by ignoring +DI/-DI, an the result was much the same. EUR/ZAR was the best pair, here ADX yielded excellent results, turning a heavily negative result into a heavily positive one.

I also wrote a bit unorthodox version of CSI (let me call it CSI' here) where I multiply ADXR, ATR (corrected as Dale describes) and the price of one pip. The last I have aqcuired from calculating the biggest expectable drawdown for VS and then scaling that to the roughly same € value for each pair (using Delta as reference), finally converting that to lots per trade and the price of one pip. I did not consider margins here just to ease calculations. I added a ranking based on this value to each of the trade that ADX selected, but haven't yet been able to find good enough correlations. It seems that even if a pair is high in this rank and ADX says trade, the trade can turn sour, and the opposite for low rankings.

Now I don't know if I should abandon ADX when looking at VS. Or maybe the other systems, which I have not studied yet, need to be combined into the package to get something reasonable out of it. I need to sort all of this out before I go live.

J.

[midulster]

Dale,

I'm sure everybody here wishes you a very happy birthday and a ton of pips because right from the first "Hi folks, I am just so excited I could 'sh1t'!" over at Parabolic SAR you've had us hooked with your honesty, encouragement, creativity and passion. Its you who has given us the gifts.

Have a great day.

Regards

Derek

PS I dug out some of my old Gary Moore, hadn't listened to it in AGES - great stuff!

[sirkeen]

Dale,

Your work on here is truly inspiring, and I cannot thank you enough for what you provide all of us.

All the 'pips are for you today dude!

Happy Birthday,

M.

[mytwopips]

HAPPY BIRTHDAY DALE!!!

May this be you best year yet, but not your best year to come.

[mastergunner99]

Just bought the book on Amazon. I'll read it from front to back and be sure to include myself in these discussions.