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

[jcc1761]

Good morning gentleman,

I just decided I'd catch up by printing some of these up so i could get away from the computer and read them with a hiliter ( and maybe a chilled beverage) under a palm tree.

Here's a copy if there are any newbies out there that want to do the same. ( I know fabio is doing something similar) this is just from 11.30.08 till today. Quite a bit of posts actually - its not all of them though. Just what jumped out at me. As i glanced thru these it was interesting to note how fast you're evolving. You guys are going to make a fortune in this business. I also picked up on the ebb and flow of enthusiasm.

Looking forward to a good uninterrupted read.

thanks all,
jcc

[fsprea]

People, what an excitement!
I’m „up to the page“ (finally!!!!) with the thread, and I managed today to “get the first glance” with the Delta platform (demo). Although I have not yet understood where the hell windows are placed when you minimize them , I managed to put Dale’s sysdicators on, and even make a bit of “cracking” (actually is just changing the color of bars to blue, given that I am a bit color blind, and blue looks to me much better than green LOL, LOL!).

Now, my aim in writing the following is by no means to “teach anything to anybody”. Rather is to learn a lot, putting myself in a trial and asking you all to verify whether I have got something out of all the work made till now with particularly the DMS. So, please, tell me all the mistakes I’m making in interpreting the system.
What anyway I did, given that this is of much concern for all at the moment, is taking out the graph that J kindly provided us in post #1703, with a lot of bad DSM trades, and see what I probably had done at that time (around Feb 08 on GBPUSD) given my understanding and interpretation of the DMS. J, if you can, please verify I’m not saying sh*t! LOL!
I know J was applying the “modified and formalized” rules for his automatic trading device, so, probably (sorry for not going through them very analytically, now, J) they are a bit different from “standard” DMS. What I’m making is applying the “standard” (besides the “blue”) Dale’s DSM Oscillator, plus a couple of “rules” I add up from Boca’s/Elder suggestions, namely: do not trade the directional system when ADX is going down; second (as also per “the old man”) take profits when ADX changes direction. I’m using the “standard” (till very recently) period of 14 days. I’m also including the “Dale’s-correct” pSAR, and I’ll comment.
So: in the first page of the included pdf there is again J’s picture. In the second page there is a screenshot of “my” Delta, GBP/USD daily, in the time period indicated by J, with the 15 trades shown by J (hopefully correct!) and an arrow indicating long (bottom up) or short (top down) positions.
Analitically:
A. Well, I would NOT have opened this positions because (i) there is no signal in my hands (no change of color of the sysdicator: no -/+DI crossing) AND (ii) ADX<20 (thick bars).

B. Again nothing. There is a signal, indeed, but ADX<20; Notice, anyway, how here, a bar before arrow “C”, there could have been a quite nice PTP short entry. And this would have been NOT whipsawed by the SAR near letter “D”, due to the extreme point rule. It could have been quite profitably closed at the SAR near “F”.

C. Again nothing for the same reason: ADX<20

D. Well: I have no “positive signal” here. BUT. Assuming I was “skipping” the first blue-to-red switch several bars before due to low ADX (and also skipping the previous PTP signal), I could have decided to enter as soon as the ADX, continuing to grow, passed the “20 threshold”. In this case, given the Extreme Point, I would probably have entered short one bar AFTER J’s D, although I would have been VERY scared by the fact that pSAR is indicating exactly the opposite. Probably I would have waited still another bar. Just some few pips until ADX changing direction would have suggested taking profits and going out the market. Anyway: few “trend” to see in this little region!

E. Really can’t understand J’s rules here. I cannot see any signal at all.

F. Also on F I have no signal. But I would have had a signal the day after. BUT: ADX is falling down! So: no trade! (not a directional one). In this region, I would recognize, also pSAR would have caused problems. The “falling ADX rule” can probably be a good filter for PTP too.

G: this could have been taken by me. Well: it’s not a wonderful one, because ADX is not very high, but also not falling. Well: let’s put like that. With “my” psychology, if I would have had nothing better to trade, that day, I would have probably tried it. A bit scared by pSAR: but this is without question ANOTHER Trading system. Again: if there was nothing better on the marketplace…

H. Again cannot see signals here. Probably J’s system has “other additional switches”.

I. Well: here I would have missed a nice trade, with my rules, because ADX is < 20, and I would have at least expected that it turned over it. There could have been a nice PTP signal, there, and, as per Extr. Pt., I would have entered long 6 bars afterwards (one bar before J’s “J”), that is one bar before “my ADX criterion” would have also allowed entering the DMS signal.

J&K: well, about entry on J see above. The thing I find quite “strange” in J's chart, is the close of his “J" long trade. Would I have entered in J, I saw no reason for closing it up three bars later. Rather, I would have followed it also “over” K, and be stopped one bar later than K by the “ADX changes direction” rule. Or eventually two bars later by PTP SAR. Anyway: some good pips in the bag!

L, M, N, O: for all of them: ADX too low. I also note that even the PTP would have been bad in this period. Although, per Extr Pt. I would have entered short only on bar N, limiting therefore the losses.

So, actually, if I am not completely mistaken here (it might well be the case!), this “terrible period” of J’s backtesting would have not looked so bad by the “conventional” DSM rules (even without changing the “standards” too much). "Not so bad" means not painful for your pockets! Surely not one where you would have filled them too! Maybe this rule of “ADX pointing upwards”, “taking profits when it turns down”, and strictly restrain from trading when ADX<20 would have saved my skin there.
Another though: it’s too early for saying that, but, you know, indicators (and sysdict’s as well) are at the bottom line instruments for better “grasping” price action. I must say that the combined visualization of DMS AND pSAR gives me quite a sort of “sixth sense” for it, at least in this situation.

A final thought: if we want to be “surer” that DSM is not going to put us in trouble, could not it be simply the case for increasing the “20 threshold”? The old man says that “as a rule of thumb” ADX value is above 20 to 25 for indicating a good trend. Instead of changing ADX periods, wouldn’t it be nice just increasing the threshold e.g. to 22 or 23? Dale, if you don’t mind I will crack again your scripts (if I MANAGE! LOL) and give it a trial.

The last question is: well, probably I would have skipped quite a lot of trades in that period on this instrument. Now: what the hell would I have traded then? (In other words: are these criteria “too restrictive” so that I would have being around idle for that period). This of course implies looking at the behavior of other instruments for that period (or other markets, or other “strange things” like OPTIONS? Maybe…).
This will be my next “exercise”.

Hopefully you did not fall asleep while reading: I know it’s by far TOO long, here! Sorry! And please: if I made “all wrong”, please, please: kick me in the a*** and tell me I must study the lesson better before speaking!!! LOL

Bye

Fabio

[dpaterso]

Good (Friday) evening everyone.

Great post Fabio and LOTS of work done!!! Well done!!!

Now:

I have had a 'cursory read' of your post (and I'm now going through your post in detail i.e. checking what YOU would have done and what parameters / rules you have used).

Just one or two points (I'll post as more things 'come to mind'):

Using ADX(14) on GBP/USD with the following parameters / rules:

DMS trades only valid if ADXR (NOTE: ADXR NOT ADX but I'll get to this shortly) above 25.

DMS trades only entered if ADX is rising.

Trades closed (whether in profit or loss) on the first downturn of ADX.

The above results in a POSSIBLE 27 trades since March 2005 and NOT ALL of them profitable. (I WILL endeavour to work out the nett result and post it though).

OK:

(This is more of a 'correction' than anything i.e. an attempt to NOT confuse).

NOWHERE does 'the old man' say that if the ADX value is above 25 etc. etc. etc. The ADXR value, on the other hand, is noted. The only time that ADX is mentioned as a 'restraint' (for want of a better word) is when he says to STOP TRADING a trend-following system when ADX goes below both of the DI lines.

Now dont' get me wrong:

I'm NOT saying that checking for ADX above 25 is a bad idea to 'filter' for DMS trades (I've not looked it yet). All I'm doing here is pointing out the difference between 'pure' Wilder and 'pure' DMS and any other interpretation thereof.

Regards,

Dale. (forexbrokersonline.net).

Edit:

As I was reading my post (as you know I DO check for grammar and spelling mistakes to the BEST of my ability) I realised that there MAY BE a misinterpretation in the line 'Trades closed (whether in profit or loss) on the first downturn of ADX'. I need to look at something regarding this and I'll post my thoughts. Sorry to confuse.

[kaalilaatikko]

Fabio, thank you very much for your post, an excellent one! It was really refreshing to go through that chart with somebody else's eyes. I guess that going through that kind of excercises you have also noticed that getting a grasp on these systems is not something that you can do with your left hand from behind your back. It is hard work. I hope that you have the inspiration to carry on!

I have some refining ideas to the system that should alleviate some of the problems that are pretty evident in my chart. Therefore I'm sorry that I'm not going to counteranalyse those trades in equivalent detail - at least yet. Some comments, though, right now for clarification:

There is a difference in our ADX charts (yours 14, mine 7), which may make the interpretation of my signals on your chart difficult.

E, H: I had ADX changing colour and DX rising.

J: The stop loss is rising closer when DX makes turns downwards. The long shadow then picks the stop, unfortunately.

C: While you are correct that entering one candle before C would have made a good trade, would you have come to the same conclusion without seeing the coming candles?


I hope that I can post something more definitive on the latest findings on the weekend. One of them is the following:

It has started to look like entering a few ticks below the low or above the high is actually not the best strategy. (Sorry, Dale, no intention to upset you with this!) What I am suggesting as replacement is to enter at market immediately after close when you get a signal. The good points in this:
- When the trade is a good one, you get more pips.
- When the trade is a bad one, you can get out with less damage.

A win-win situation, not too bad! But as always in this business, the coin has two sides. Here the opposite side contains some bad trades that you could avoid by placing a stop order. But it seems that the benefits overweight the drawbacks.

BR,
J.

[fsprea]

Both of you: thank you, thank you, thank you!!!!

Just, again, some "cursory thoughts". J, first.
E & H (and the other signal I was "missing"): well it is now all clear with the difference in ADX(periods). Unfortunately your E&H were some of your "unlucky". Would it somehow suggest that in this precise instance the 14 could work better than the 7?
C: actually what I mentioned as a possibility is NOT DSM but PTPS. As far as I have understood Dale's explanation of PTP (coming from a DEEP Nobel-Price grade in pSAR"ology" ), I would have had a short signal less or more where "B" is located, but, given the low of this same bar, the limit order (some pips below), would have been crossed just a bar before C. So: it was not a "retrospective" decision. Well, you know, as far as it is possible after knowing the story already, as in all "visual" backtesting!!!!

Dale: great! Thanks. Really, I was so enthusiastic today that I started mixing up with ADX and ADXR. Probably it's just a matter of chance, but in the present chart all but (possibly D and I), E, F, G, H, would also fall in times of ADXR<20. I must also say that, now, with this correction by yours, I'm anyway quite surprised by the "positive effect" of (wrongly) waiting for ADX(without R) coming up above 20! Anyway, at least the direction of ADX (upwards instead than downwards) seems to me an interesting issue.

But then, let me also go a bit further with another subject:

The above results in a POSSIBLE 27 trades since March 2005 and NOT ALL of them profitable. (I WILL endeavour to work out the nett result and post it though).

First: looking forward with HUGE interest to your results... but

MAN! WHAT DO YOU WANT MORE THAN THAT??? lol LOL

Sorry: let's just make a bit of "finger math".

March 2005 to date is roughly 46 months. With 22 days/month this sums up to 1012 days (daily candles). Let's say that on average one DMS trade lasts 15 days (include also the "stellar" downtrends you would have liked to follow "entirely" for thousands of pips in the past months... two-three months-long trades?). 27 trades / 15 days long = 405 days. This means that, on this single instrument, with the sort of "safe" DMS (if this is the case) you're on the market 40% of the time. How were the statistics about trending/ranging?

Then you have let's say 20 "good" forex pairs to look at; some commodity (3?), let's also say 100 good stocks (just pick your preferred with enough liquidity), ten ETFs and probably much more of the like (I think that, also for YOU it could be difficult keep an eye on more than that: but I can really underestimate you LOL ). Just make 20 trades (not 27!) in 46 months for all this stuff: you have a potential of 133*20 = 2660 trades in the same period: 57/month; 15/week. Dale, my friend, are you still hungry? Still feel "unemployed"? LOL! Just take the 20 fx pairs (or 20 instruments) a "more common mortal being" is able/willing to follow: 400 trades, 8/9 trades per month. Few? It's already enough to be forced to lower the single position margin not to overflow the total capital invested according to "our" MM rules!

And this is just by looking at a "trending mode" system. Think about the poor guys like me that were coming to visit you at forex just because, unfortunately there was just too much movement around these times for going on playing (and making $) on things that are moving as few/slowly as possible!

I mean: what do you want to get out from the DAILY timeframe? And, by the way, I'm fully WITH you in excluding (at least in my current attitudes) switching to shorter.

Well: don't take it toooo seriously. But I wanted just to make a point here. I think the number of trades you presented above is not surprising, rather quite reasonable. It's not scaring me at all. Profit potential with THAT "LIMITED" number of trades tends to infinite. Stretching too much systems to squeeze out more can produce wonders, but also disasters. In finance the higher the reward the higher the risk: I think we should always remember this "Newton Law".

Bye

Fabio

[kaalilaatikko]

Good morning! Just a short notice that I have not faded out, as I promised to try to post something about my findings during the weekend. This ADX+DX combination really seems very promising, and I have done a lot of backtesting for it by varying different factors therein. I'd still like to hold these results a bit as I haven't had enough time to have a look at all those things I've wanted to.

Sorry for being a bit short-worded now, but if I were Akram, I might say that I might be about to dance soon!

J.

PS. It seems again to me that entering by stop orders is the way to go, after all.

[midulster]

I know its only one day's worth of data and I havent been taking all the signals but does anyone else have a strong link between their current green positions and the highest (not just >25) ADXRs?
Eur/gbp 46
Usd/jpy 42
Gbp/chf 37
Gbp/jpy 37
Eur/cad 35
Eur/nzd 35
Eur/aud 33
Eur/usd 33
Nzd/chf 32
Usd/chf 32
Chf/jpy 31
Nzd/usd 28
Gbp/cad 27
Cad/jpy 26
Nzd/jpy 26
Cad/chf 25
Gbp/aud 25
Audjpy 24
Eur/chf 24
Gbp/nzd 24
Aud/chf 23
Nzd/cad 22
Gbp/usd 19
Aud/nzd 18
Usd/cad 18
Aud/usd 17
Aud/cad 12

Derek

[dpaterso]

Good (Monday) morning all.

'Father Ted':

I WISH!!!

I have a VERY STRONG link between 100% of my currently open positions and the color RED!!! LOL!!! I cannot therefore comment on your question!!! Would I be correct in ASSUMING, though, that YOU DO in fact have a strong link between the highest ADXR values and the color GREEN???

Regards,

Dale. (forexbrokersonline.net).

[midulster]

Yeah - 4 green positions all ADXR >33
4 red positions 3 with ADXR <25

I didn't choose them based on ADXR - just thought I'd add a note from Prof. Hindsight. Will check when I get home from work.

Derek

[dpaterso]

Point taken 'Prof. Hindshight'!!!

That being said though: more often that not and I MEAN MORE OFTEN THAN NOT waiting for ADXR > 'anything other than zero' costs you LOADS of points / pips from what I can see so me: as long as ADX is not below BOTH the Di's and I get a signal I'll place the order.

I MUST NOTE however: the thread is REAL QUIET AGAIN!!! LOL!!! ANY BETS as to WHY??? Could it be that THIS is one of THOSE Monday's to which I have become accustomed to i.e. those INFURIATING Monday's where 'the works' goes against your every open position??? LOL!!! Once again: MONEY MANAGEMENT IS KEY!!!

But NOT to worry everyone (Akram and Cody: your silence is noted by the way): there is M-U-C-H work going on 'behind the scenes' as the moment so just 'hang in there'!!!

Regards,

Dale. (forexbrokersonline.net).