Spydertrader's Jack Hershey Equities Journal

[svrz]

Quote from g33m4k:

Moving right along, here is version 19. Not trying to step on any toes. Just including what svrz has already coded up in WL into our ever expanding workbook. Just to address any considerations, all the data works into the "Easy Scan" portion of the QA-ROI.doc. Several posts across a few threads specifically explain what's happening here. This may require some light reading for this to settle in. This should allow for greater range in tracking the progression of equities across it's cycle. If I am not mistaken, staging is used to isolate which cycle quadrant an equity is currently in as it rotates into and out of play for us. Can anyone elaborate on this?

If I remember the discussions correctly since 1999, I don't think staging works very smoothly. David Marshall made some great attempts in making a science out of this but to no avail -apparently.

One item does seam to be in error and that is the "Stage 1 or Stage 3" filter. It is not possible for the LO 5 days ago to be 10% higer than the HI 30 days AND also for the LO 30 days ago to be 10% higher than the HI 5 days ago... Is the second inequality sign backwards?

Believe it or not, I don't have access to Excel at the moment. Check out the correct version of the equations by David:

http://tinyurl.com/5yezc

I think these are the equations:

AAA Stage 1 or Stage 3

(H30 < 1.1 * L5) AND (H5 < 1.1 * L30)


AAA Stage 2 or Stage 4

((MAXC126 - MINC126) / MINC126) * 100

Take care and thanks for all the great contributions that you are making.

 

[g33m4k]

Quote from svrz:

If I remember the discussions correctly since 1999, I don't think staging works very smoothly. David Marshall made some great attempts in making a science out of this but to no avail -apparently.



Believe it or not, I don't have access to Excel at the moment. Check out the correct version of the equations by David:

http://tinyurl.com/5yezc

I think these are the equations:

AAA Stage 1 or Stage 3

(H30 < 1.1 * L5) AND (H5 < 1.1 * L30)


AAA Stage 2 or Stage 4

((MAXC126 - MINC126) / MINC126) * 100

Take care and thanks for all the great contributions that you are making.

NO PROBLEM . Just trying to keep up with you . It's my way of learning.

About these equations... The Stage1/3 equation does not make logical sense. The result here will ALWAYS result in an empty list. Interpreting this equation, it states that the names from our universe that are in Stage2/4 is determined by whether or not an equity's historical LO 5 days ago was at least 10% higher than the HI 30 days ago and additionally the equity must have had a HI 5 days ago that is no more than 90.9% of its LO 30 days ago.... No such pair of HI/LO ranges for any set of days can ever satisfy this condition... If we can determine by what is meant by "Stage" 2/4, we can correct this equation.

Best Regards,
G33M4K

PS...
Just for precision because I'm pretty anal
about this stuff, just want to state that
TC2000 code is such that H=H0 = present day HI,
H1=1 day ago, ... H30=HI 30 days ago

 

[gallas2]

Quote from g33m4k:

NO PROBLEM . Just trying to keep up with you . It's my way of learning.

PS...
Just for precision because I'm pretty anal
about this stuff, just want to state that
TC2000 code is such that H=H0 = present day HI,
H1=1 day ago, ... H30=HI 30 days ago

IMO, trying to over-engineer this approach doesn't work. I simply attempt to eyeball the 'essence' of the approach (ie pullback, du, frv). If it doesn't fit precisely with a formula it doesn't mean you discard it.

DM and g33m4k have made excellent contributions - no knock against - it is simply their background that makes them want to understand everything to the finest detail. I am more concerned with making money than finding the perfect candidate.

If you must engineer something - engineer the money management side of things -- this will keep you out of trouble.

 

[g33m4k]

Quote from gallas2:

...
If you must engineer something - engineer the money management side of things -- this will keep you out of trouble.

Fair Enough... I just would like to state that what may be seen as money management, I view as quality assurance. Jack made the following comments regarding money management...

http://groups.google.com/groups?q="...elm=01be1b50$55f4e900$413384ce@default&rnum=2

Of course, I do not take everything Jack posts to be the end all, however I do believe we are looking at two sides of the same coin. Money Management/OverEngineering/Quality Assurance are all the same thing. Since I don't trust myself to catch all the flaws since I am human, I program those flaws which can be defined programmatically. Resorting to the race analogy, I want nearly perfectly tuned vehicles which will not stall at any time during it's up cycle. I believe the workbook is providing an extremely thorough systematic check of our vehicle to determine the vehicle's reliability. Ideally, a perfectly reliable vehicle does not need a bumper to bumper warranty (money management), since it has proven consistently and checked out to be reliable. Because of my quality criteria, I am performing as rigorous a test as is mechanically possible to ensure that nothing indicates that the vehicle will stall prematurely... However, I see that I have overkilled the programming and thus will need to resort to another means for trying to learn the mathematical rigors I believe to be relevent and meaningful...

G33M4K

 

[g33m4k]

Quote from svrz:

If I remember the discussions correctly since 1999, I don't think staging works very smoothly. David Marshall made some great attempts in making a science out of this but to no avail -apparently.

...
I think these are the equations:

AAA Stage 1 or Stage 3

(H30 < 1.1 * L5) AND (H5 < 1.1 * L30)


AAA Stage 2 or Stage 4

((MAXC126 - MINC126) / MINC126) * 100

...

svrz...

After sleeping on it , the AND is supposed to be an OR. The first equation identifies Up/Down quadrants of the cycle, the second focuses on the extremes of the cycle (peaks/troughs).


AAA Stage 1 or Stage 3

(H30 < 1.1 * L5) AND (H5 < 1.1 * L30) => (H30 < 1.1 * L5) OR (H5 < 1.1 * L30)


G33M4K the Newb

PS.
BTW, these are Jack's TC2000
equations, not DM's. Checking
the context of the original thread
shows that DM was given these
equations, and he posted them
so that he may share and ask
Jack questions about them.

 

[svrz]

Excellent job as usual, G33M4K!

Do you have access to WL 3.0? I think you can download a 30 day trial and if you e-mail the company, I think they will give you a 60 day license that will completely unlock the trial version. If you are interested, I can post the scans for these equations so you or anyone else can automate the process.

Take care

 

[gallas2]

Quote from g33m4k:

Fair Enough... I just would like to state that what may be seen as money management, I view as quality assurance. Jack made the following comments regarding money management...

http://groups.google.com/groups?q="...elm=01be1b50$55f4e900$413384ce@default&rnum=2

Of course, I do not take everything Jack posts to be the end all, however I do believe we are looking at two sides of the same coin. Money Management/OverEngineering/Quality Assurance are all the same thing. Since I don't trust myself to catch all the flaws since I am human, I program those flaws which can be defined programmatically. Resorting to the race analogy, I want nearly perfectly tuned vehicles which will not stall at any time during it's up cycle. I believe the workbook is providing an extremely thorough systematic check of our vehicle to determine the vehicle's reliability. Ideally, a perfectly reliable vehicle does not need a bumper to bumper warranty (money management), since it has proven consistently and checked out to be reliable. Because of my quality criteria, I am performing as rigorous a test as is mechanically possible to ensure that nothing indicates that the vehicle will stall prematurely... However, I see that I have overkilled the programming and thus will need to resort to another means for trying to learn the mathematical rigors I believe to be relevent and meaningful...

G33M4K

I like the race analogy. Expanding on that, you can't win the ragatta if you are still at the dock screwing with the rudder. IE trade the approach with real money and it will re-inforce what a good candiate looks like real quick.

Re: money mangement - my account likes it. JH has an inheren MM approach by virtue of the "streams" and stop losses.

 

[g33m4k]

Quote from gallas2:

I like the race analogy. Expanding on that, you can't win the ragatta if you are still at the dock screwing with the rudder. IE trade the approach with real money and it will re-inforce what a good candiate looks like real quick.

Re: money management - my account likes it. JH has an inheren MM approach by virtue of the "streams" and stop losses.

My perspective on the race is slightly different. I am in no shape or form building my own vehicle and thus rudders do not fit into my perception. I am simply building a comprehensive diagnostic tool fit for the vehicles built by the employees of the companies meeting our quality criteria. Any diagnostic device with improperly tuned settings will provide faulty read outs which for us is the ultimate cost since we cannot recover the time spent in a vehicle that has stalled and even worst, has gone into reverse. My intention is to tune my diagnostic to spot the top currently performing and fastest equity vehicle in this race, and then jump into these vehicles (current leader board) until my diagnostics device informs me of more opportune vehicle(s) to jump in to (rotating streams). This ensures that I am always winning and evidently is SCT of equities, this realization is frightening since it is a consequence of the approach that I am not prepared for. Believe it or not, the potential here is possibly greater than futures which has few vehicles. My crutch for now is that I am seeing at least 2 or 3 intermediate plateaus without having worked thru the mechanisms for beginners. I speak for myself when I say that understanding the whys makes everything KISS... So, at the encouragement of others, I will continue to churn .

G33M4K the Newb

 

[svrz]

Quote from g33m4k:

My perspective on the race is slightly different. I am in no shape or form building my own vehicle and thus rudders do not fit into my perception. I am simply building a comprehensive diagnostic tool fit for the vehicles built by the employees of the companies meeting our quality criteria. Any diagnostic device with improperly tuned settings will provide faulty read outs which for us is the ultimate cost since we cannot recover the time spent in a vehicle that has stalled and even worst, has gone into reverse. My intention is to tune my diagnostic to spot the top currently performing and fastest equity vehicle in this race, and then jump into these vehicles (current leader board) until my diagnostics device informs me of more opportune vehicle(s) to jump in to (rotating streams). This ensures that I am always winning and evidently is SCT of equities, this realization is frightening since it is a consequence of the approach that I am not prepared for. Believe it or not, the potential here is possibly greater than futures which has few vehicles. My crutch for now is that I am seeing at least 2 or 3 intermediate plateaus without having worked thru the mechanisms for beginners. I speak for myself when I say that understanding the whys makes everything KISS... So, at the encouragement of others, I will continue to churn .

G33M4K the Newb

This is exactly how I feel. Diagnostics is the appropiate term to apply here. In essence, we need to have the most modern and automated shop in town.

I think WL is the perfect platform for our needs. We can automate this process from start to finish. Well, there is one manual step and that is to get the list from Stockcharts.com into WL.

 

[makosgu]

Quote from g33m4k:

AAA Stage 1 or Stage 3

(H30 < 1.1 * L5) AND (H5 < 1.1 * L30) => (H30 < 1.1 * L5) OR (H5 < 1.1 * L30)


G33M4K the Newb

Apparently I hadn't gotten enough sleep and therefore I am posting to clarify. It is entirely possible to produce a list with the initial Stage 1/3 equation. Here's an Example...

HI/LO 30 DAYS BACK = 32/29
HI/LO 5 DAYS BACK = 31/30

THUS, using the original equation would provide a list of stocks that seam to have a relatively tight/close price range when contrasting the range of 5 and 30 days back. So I am left back where I was, should the AND be the more intuitive OR, or does the original equation take advantage of a highly correlated and unapparent bar signal (some sort of trend reversal)??? I would attempt to run a regression but I'd much rather not. Can anybody or better yet, has anyone been using the tc2000 eqn properly? My apologies for lack of broader context understanding...


AAA Stage 1 or Stage 3

(H30 < 1.1 * L5) AND (H5 < 1.1 * L30) [is this condition part of all trends]
(H30 < 1.1 * L5) OR (H5 < 1.1 * L30) [1-more often an uptrend; 2-more often a downtrend]


Unfortunately, my universe is small so it's difficult to watch names move from equation list to equation list. Does anyone have a formidable Final Universe list that they care to share that is larger than Spyder's? Otherwise, I assume I will begin the long haul, binders and all.

In any event, in attempting to identify a comparable derivatives function, I fell into a posting regarding MACD's as derivatives. I know most people in ET fight against long posts so my recommendation would be to print this out and leave it as bathroom reading.

If you don't like healthy technical docs, avoid the link.
http://groups.google.com/groups?hl=en&lr=&selm=8d2vov$rgo@flatland.dimensional.com&rnum=1


Just out of curiosity, is anyone here currently employing Elliott Waves? I toyed with it today and applied it to INTC and much to my surprise, it is one of those favorably bizarre curiosities when applying the pseudo rules. I want to attempt some LINEAR regressions but for the time being, I am focusing on taking the plunge some time next week. I'd much rather be in a strong bear market for plunging since I assume it would make journalling and debriefing alot more interesting with regards to various aspects of the vehicles under the harshest of elements.

For those who may be interested, I am attaching v0.20. It has several corrections regarding the tc2000 equations (tc2000 vol constants have an implicit factor of 100 embedded in their value).

gallas2, if I may request, I would greatly appreciate if you could provide/post an anotated chart of an equity that I could learn from so that I can focus on what needs to be eyeballed. As you had covincingly mentioned, there's only so much than can be quantified. I am quickly spending time that can be better used for capital appreciation as I am now anticipating taking the plunge.

Regards & NJOY
G33M4K - Newb