Question for Grob/Hershey...

[Trader666]

Quote from easyrider:

I have yet to meet one backtester who is making any money in the market. If you are or someone else is and can back it up I might have to change my opinions.

P.S. For those who subjectively dismiss objective backtest results, not only without trying to replicate them but without even knowing how, I leave the following thought from the late Carl Sagan:

"In science it often happens that scientists say, 'You know that's a really good argument; my position is mistaken,' and then they actually change their minds and you never hear that old view from them again. They really do it. It doesn't happen as often as it should, because scientists are human and change is sometimes painful. But it happens every day. I cannot recall the last time something like that happened in politics or religion."

 

[easyrider]

I think that backtesting probably did work a few years ago, before everybody and their brother climbed on board. When you have umpteen thousand people looking at the same thing where is there room for an edge? You seem like a serious guy and if you can find a way to make it work for you then more power to you. I need to find something to do on the weekends. Since I gave up fishing all I do is end up in these debates that go nowhere.

 

[makosgu]

Quote from Trader666:

Hi MAK,

Of course I used historical fundamental data! Thought I explained that but must not have been clear. I used only data that was available at the time of each test period -- historical technical and historical fundamental -- and tested longs only. First (per the criteria Jack originally listed, see below) without the float and other stuff and then later with it. As for the intraday example you gave, Jack's description of rockets that I tested was for daily charts:

"To find a rocket you only have to have your quality universe (they are very high quality stocks with EPS and RS at 80 to 90 percentiles each. The only indicator that is significant is the Stochastic (5,3,3). What you look for on your daily charts are the stochastic rising deliberately to the 80% line and overshooting it and then critically damping on the line; perturbations will cause it to look entwined on the line. There you have it."

I think you're way more of a discretionary trader than I am. I prefer to apply the minimal discretion that I do use to stuff that does backtest well. Please see my reply to easyrider...

Trader666

Fair enough. I respect your capabilities as a programmer. I understand the backtest components and the probability space that it deals with. Personally, I have had to run monte carlo simulations of sample sets to estimate Std Dev, mu, and confidence intervals to see where a particular set of go/no go's winds up in the space of Win/Loss. There are a bunch of very fine folks who know how to deal with the probability space and accelerate the belief component. Acrary is a pro at mapping probability to profitability. You would definitely like his stuff...

I believe you understand the meat of the components and perhaps need only a few tweaks to weed out some of the no go rockets...

Just one other thing about the STOCH since it is not often mentioned. The STOCH is a relative indicator while the MACD is an absolute indicator. As a pair they are complementary and I will try and be specific. SO the settings on the 2 indicators are configured to be in synch with one another. When you look at the formula of the STOCH, you find that it's design is to indicate where the current price is in relation to the last x number of bars. As a result, >80 reading is saying that the current price is extending beyond the range of the last x number of bars. Thus, the hold as long as bars are continuing to extend the range of the last x number of bars. However, when the range of the last x number of bars is relatively low if not flat, the indicators will still indicate where the current price is wrt to the last x number of bars. Hence, the need for the absolute indicator MACD. The MACD is simply a direct non lagging parrallel of the price action. If you overlayed a MACD over price, you would see that it matches the price movements. For some reason, folks use settings in which there is an offset between the price movement and their fast line. That all just sums up to lagging which I would think is undesirable. The MACD takes care of when not to gate the STOCH signals.

Best Regards,
MAK

 

[Spydertrader]

Quote from Trader666:

For those who subjectively dismiss objective backtest results,

1. How did you obtain the historically accurate RS and EPS Ranking? Could you please provide a link to the data provider, or describe how you created the rankings on your own.

2. I see no mention of using equities which "Cycle Five Times in Six Months for 20% Gains occurring over 6-8 days" as mentioned in this post.

3. Which time frame and which equities did you test? Perhaps, others you have accused of subjectively dismissing your efforts might benefit from you sharing your testing methods. My own backtested (and real world trading) results show significantly different results than you have posted here.

Your claim to "only backtest specific criteria posted by Jack" doesn't fly when others in the same thread provided additional, and pertinent detail which you chose to ignore. To now claim people are adding conditions after the fact is laughable. It's one thing to ignore important information, and quite another, to claim nobody provided it to you in the first place.

In other words, you appear to have proven a trading method doesn't work. The fact that nobody in either discussion group (futures or equities) currently trades the method you tested seems to have been missed by you. Thank you for your contribution to the effort.

Good Trading to you all.

- Spydertrader

 

[Trader666]

Quote from easyrider:

I think that backtesting probably did work a few years ago, before everybody and their brother climbed on board. When you have umpteen thousand people looking at the same thing where is there room for an edge? You seem like a serious guy and if you can find a way to make it work for you then more power to you. I need to find something to do on the weekends. Since I gave up fishing all I do is end up in these debates that go nowhere.

Wow, so many contradictions! I pointed you to a whole site of real-money, real-time funds, many of which are traded mechanically, and you still doubt it? You wrote:

"When you have umpteen thousand people looking at the same thing where is there room for an edge?"

First of all, who said all mechanical traders are looking at the same thing? There are an infinite number of possible mechanical trading systems. Second, the day before yesterday you wrote: link

"I am a pattern trader. I do not see how this kind of trading can ever go away. A bearflag in a strong downtrend is a moneymaker and barring some drastic change in human nature it always will be."

So there aren't "umpteen thousand people" looking at bear flags? See any contradiction? Your post was a joke, right?

 

[Trader666]

Quote from makosgu:

Fair enough. I respect your capabilities as a programmer. I understand the backtest components and the probability space that it deals with. Personally, I have had to run monte carlo simulations of sample sets to estimate Std Dev, mu, and confidence intervals to see where a particular set of go/no go's winds up in the space of Win/Loss. There are a bunch of very fine folks who know how to deal with the probability space and accelerate the belief component. Acrary is a pro at mapping probability to profitability. You would definitely like his stuff...

I believe you understand the meat of the components and perhaps need only a few tweaks to weed out some of the no go rockets...

Just one other thing about the STOCH since it is not often mentioned. The STOCH is a relative indicator while the MACD is an absolute indicator. As a pair they are complementary and I will try and be specific. SO the settings on the 2 indicators are configured to be in synch with one another. When you look at the formula of the STOCH, you find that it's design is to indicate where the current price is in relation to the last x number of bars. As a result, >80 reading is saying that the current price is extending beyond the range of the last x number of bars. Thus, the hold as long as bars are continuing to extend the range of the last x number of bars. However, when the range of the last x number of bars is relatively low if not flat, the indicators will still indicate where the current price is wrt to the last x number of bars. Hence, the need for the absolute indicator MACD. The MACD is simply a direct non lagging parrallel of the price action. If you overlayed a MACD over price, you would see that it matches the price movements. For some reason, folks use settings in which there is an offset between the price movement and their fast line. That all just sums up to lagging which I would think is undesirable. The MACD takes care of when not to gate the STOCH signals.

Best Regards,
MAK

Hi MAK,

Thanks for actually putting some thought into your reply I've been a practitioner of MCS for many years myself and do like Acrary's stuff.

Agree with your point on MACD and Stoch... I've also found that complementary indicators can be tuned to yield something together that neither alone can provide.

Trader666

 

[easyrider]

Reading back over my comments while Im sober I can see some of them were out of line. Par for the course on Saturday afternoons, for me. I seem to go into a state of irrational exhuberence when the week is over. I dont have the link but I also posted some time back that any posts by me on Saturday should be ignored.

 

[Trader666]

Quote from Spydertrader:

1. How did you obtain the historically accurate RS and EPS Ranking? Could you please provide a link to the data provider, or describe how you created the rankings on your own.

2. I see no mention of using equities which "Cycle Five Times in Six Months for 20% Gains occurring over 6-8 days" as mentioned in this post.

3. Which time frame and which equities did you test? Perhaps, others you have accused of subjectively dismissing your efforts might benefit from you sharing your testing methods. My own backtested (and real world trading) results show significantly different results than you have posted here.

Your claim to "only backtest specific criteria posted by Jack" doesn't fly when others in the same thread provided additional, and pertinent detail which you chose to ignore. To now claim people are adding conditions after the fact is laughable. It's one thing to ignore important information, and quite another, to claim nobody provided it to you in the first place.

In other words, you appear to have proven a trading method doesn't work. The fact that nobody in either discussion group (futures or equities) currently trades the method you tested seems to have been missed by you. Thank you for your contribution to the effort.

Good Trading to you all.

- Spydertrader

Spydertrader,

I'm not usually one to knock someone who appears to be slugging it out in the trenches. I say appears because a very quick look at your journal shows at least one trade taking place at a price the stock didn't trade at that day. Anyway, your response seems more like something from the propaganda minister of a cult than something a trader would write. First you fabricated a quote, then you totally misrepresented what I wrote, and finally you closed with "Thank you for your contribution to the effort." "The effort?" LOL. Now that I think about it, there are plenty of parallels between Grobians and Woodies followers. But that could be a whole new thread A few quick points:

1. I never claimed to "only backtest specific criteria posted by Jack" and I didn't write those words. You did. What I did say I would do (and it's in the post you linked to) is: "test "rockets" first with just EPS & RS ranks and Stoch" (which are the three elements of a "rocket" from Jack's post) and then "add in the other stuff you mentioned (price, average volume, float, etc)." Which is exactly what I did: link So please spare me the history rewrite.

2. You wrote: "you appear to have proven a trading method doesn't work. The fact that nobody in either discussion group (futures or equities) currently trades the method you tested seems to have been missed by you." First, that's exactly what I told you I was going to backtest here, so why didn't you make that point then? Second, why isn't anyone trading it? That method is straight from the horse's mouth. You're not saying that Jack actually put out something that doesn't work, are you?

3. To answer your question about how I got my rankings, I use a proprietary database. But here's a resource that anyone can use to get historically accurate rankings of stocks, based on any of hundreds of fundamental criteria: http://www.portfolio123.com I think they even have a free 30 day trial. Stocktables.com is primitive in comparison. But just as a valid, profitable technical trading system will show roughly similar (but not always identical) results whether the data is from eSignal, CSI, TC 2000, etc., the same is true of fundamental data from different providers. So don't split hairs when the rankings don't match exactly. Because they won't. But if the method is any good, it won't matter. Things that really work in the markets don't fall apart when you change tiny details. Warning: don't check this site out if you don't want a reality check on Grobian Trading Theory (GTT).

 

[Spydertrader]

Quote from Trader666:

Spydertrader,

I'm not usually one to knock someone who appears to be slugging it out in the trenches. I say appears because a very quick look at your journal shows at least one trade taking place at a price the stock didn't trade at that day.

I recall the day specifically. As explained in this post, my broker (at the time) did me a favor (as pay back for a previous error) with an internal fill. Of course, you remain free to believe whatever you like.

Quote from Trader666:

Anyway, your response seems more like something from the propaganda minister of a cult than something a trader would write. First you fabricated a quote, then you totally misrepresented what I wrote, and finally you closed with "Thank you for your contribution to the effort." "The effort?" LOL. Now that I think about it, there are plenty of parallels between Grobians and Woodies followers. But that could be a whole new thread A few quick points:

Placing the quotation marks around the sentence fragment to imply those were your exact words is an error on my part. Certainly, you didn't use the exact phrase. However, here is what you did say:

Quote from Trader666:

1. I never claimed to "only backtest specific criteria posted by Jack" and I didn't write those words. You did. What I did say I would do (and it's in the post you linked to) is: "test "rockets" first with just EPS & RS ranks and Stoch" (which are the three elements of a "rocket" from Jack's post) and then "add in the other stuff you mentioned (price, average volume, float, etc)." Which is exactly what I did: link So please spare me the history rewrite.

Aren't these your words?

Quote from Trader666:

I tested "rockets" because specific criteria for them were posted and because the concept seems to make a lot of sense...

And again, you quoted Jack's words here.

Quote from Trader666:

It is within the above framework that I've evaluated elements of your "teachings." For example, in a rare moment of clarity (Fri, Jun 21 2002 12:57am in misc.invest.technical) you described "rockets":

Now, since we agree you didn't use the exact phrasing, perhaps you could clarify: Who's 'specific criteria' and 'teachings' did you test / evaluate again?

Quote from Trader666:

2. You wrote: "you appear to have proven a trading method doesn't work. The fact that nobody in either discussion group (futures or equities) currently trades the method you tested seems to have been missed by you." First, that's exactly what I told you I was going to backtest here, so why didn't you make that point then? Second, why isn't anyone trading it? That method is straight from the horse's mouth. You're not saying that Jack actually put out something that doesn't work, are you?

I did tell you. What is more, I linked to the exact methods used in Journal One. In your reply you indicated how certain aspects of your tests appeared "self-evident" to you. Since you appeared to already 'know' everything, I saw no reason to continue to provide you with additional clarification. Since people who trade the methods outlined in Journal One use a Universe of Stocks which cycle a minimum of five times in six months for 20% (or greater gains) over a period of six to eight days, and since your tests failed to account for 'cycling' stocks, it should therefore be self-evident to you why nobody trading equities trades the methods you tested. In addition, you have posted your findings to the 'Futures Trading Discussion' Group. Since this discussion trades Index Futures, and you tested equities, it should therefore be self-evident to you why nobody trading futures trades the methods you tested.

Quote from Trader666:

3. To answer your question about how I got my rankings, I use a proprietary database. But here's a resource that anyone can use to get historically accurate rankings of stocks, based on any of hundreds of fundamental criteria: http://www.portfolio123.com I think they even have a free 30 day trial. Stocktables.com is primitive in comparison. But just as a valid, profitable technical trading system will show roughly similar (but not always identical) results whether the data is from eSignal, CSI, TC 2000, etc., the same is true of fundamental data from different providers. So don't split hairs when the rankings don't match exactly. Because they won't. But if the method is any good, it won't matter.

Thank you for providing the link to http://www.portfolio123.com. Although the 30 day free trial limits the usefulness of the web site, I do plan to test it out thoroughly in the future.

Quote from Trader666:

Things that really work in the markets don't fall apart when you change tiny details. Warning: don't check this site out if you don't want a reality check on Grobian Trading Theory (GTT).

Yet, you left out one 'tiny detail' in your tests (cycling stocks), and it appears to have thrown your results out of whack.

Good Trading to You.

- Spydertrader

 

[easyrider]

http://funnieststuff.net/viewmovie....wmv&path=/b/bumpinspeakers/bumpinspeakers.wmv

Just to lighten the mood.