The Stochastic Indicator

[jack hershey]

Quote from tampa:

With only a very basic understanding of the method, and understanding that Jack may well have passed on some trades, I calculate a gain of 9 1/2 points on the day. This was achieved simply by taking all of the 14-1-3 signals on a five-minute chart.

On the other hand, passing on about a 1/3 of the signals, and bailing early because of my own use of support/resistance, plus some second-gussing, and some disbelief, I walked with 5 1/4 points on the day.

nether total is bad...

Yhe 18 th and 19th were flat. Today, you did have some trending. To pull down 5 points or 150 bucks on one contract for 250 day a year is only 60,000 dollares plus not compounded. If you compound it is over 90k. Soon we will be getting making money understood.

thanksso much for stating your feeling and relative position to what's possible.

Both the dawg scenario and yours are very helpful to others.

 

[jack hershey]

Read the Journal and read the grouping of people taking part. they will form great examples of changes in level of performance.

If you have William J . O'Neil's book "How to Make Money in Stocks" set up a log of where you have been in increments of years of your choosing to see when and where you lost the 18 common mistakes most people make as noted in chapter 20.

Make a chart and form columns of the past and leave some space for the future.

This past week we had two flat days and then Friday which came close to flat. a week ago on Friday there were four nice 10 plus point trends.

We can get going on the beginner stuff and with blueberry cake keeping tabs on how each improvement we make, we can see how good it can get.

One thing that sometimes happens is "intuition" and wisdom" sets in. Who knows what happens when people get creative and proactive.

If you look at the journal and indent it so you can more easily see he order of importance of things, then you can look at the big chuncks first instead of being myopic.

If you get to seeing sequences, for example you will begin to see the market turn to slow motion. this also lets you catch on to the fact that by choosing signals from indicators properly, you then leave the premise that indicators are always lagging.

Take the derivative of the fast line of the MACD and note that when it peaks and begins to lessen you have just seen a leading indicator signal.

Do not use the design default of 12, 29, whatever for MACD, that went out the window when PC's were invented. It bridges trading cycles in todays investor approaches and algorithms and paradigms.

Plan on mixed markets, flat markets and good trending markets next week.

I am going to stay behind the beat for a couple of weeks to find out who can learn what first,

Then I will post 5 to 15 minutes ahead of the market for a while so you can get used to preparing properly for a good shot at some trends.

My health has picked up to the point that I will be on the road about three days a week until early August according to my calendar. This stuff will be in software by the end of that period.

 

[bdixon619]

Quote from jack hershey:



Yhe 18 th and 19th were flat. Today, you did have some trending. To pull down 5 points or 150 bucks on one contract for 250 day a year is only 60,000 dollares plus not compounded.

jack hershey,

check this out. multiply the equity times 5% each day and you know what it costs you in IB commissions and how many points per trade and how many contracts and how many round turns to achieve it in 21 days.

Bruce

 

[kempo]

What is meant by?

Set ES to 10 point scaling

and

What is meant by... "always (last one in)" and "always (first one out)

Enter as both show on the 20 (short) or 80 (long) always (last one in)
Exit as both leave the 20 (short) or 80 (long) always (first one out)

Also, are the settings for stoch & macd, simple or exponential?

Thanks!

kempo

 

[jack hershey]

Thanks for the questions. Many people will have the same items to deal with.

Quote from kempo:

What is meant by?

Set ES to 10 point scaling
**** Short answer. Look at the labels on your price graph. Make them read in a 10 point series. something like 830, 840 ,850.

Long answer: The software people are not traders. They are selling a product to you that they want to be appealing. They see "filling the screen" as more important than having a default that makes you money. To focus on market characteriestics you need a consistent viewpoint. You can better acheive this with a constant scaling deliniation. The biggest ten point illustration is achieved by scaling up until a new scaling shows, then backing down so it just disappears again. In this way you have a consistent 10 point scale and the screen may not be filled. If we get some all day trends (like on news that the war has been cancelled or the war is beginning) then to stay on the screen we probaly will have a 20 point or 50 point scale.

"Greenspans" occur occasionally too. A standard "Greenspan" is 150 points both ways on the DJXX in an hour and 1/2. He goes off on tuesdays or wednesdays about 7 or 8 minutes before his announced timing of 14:15.

We will get into scaling etc as we go further into fractal pairs as anticipation which is the replacement for prediction and backtesting (back testing is a self fulfilling method of backwards predicting)


What is meant by... "always (last one in)" and "always (first one out)

**** Short answer. There are two lines on the stochastic. One precedesthe other. "last one in" means both lines must be there so the signal is the slow line. Verse visa, "first one out" means the signal is when the fast line leaves. Watch Trendy if she posts. She may get into advanced cheating by just operating on the fast line and ingnoring the slow line.

Cheating will be a sport here and it is different than dumbness. cheating is one way I lead into anticipation but then to perfect it we use fractal pairs which then actually stops the cheating aspect. we have a phyicist here too, I will let him use different leading indicators by applying the equivalent of differential calculus to non-continuous functions. You can make up a way of using statistics to simulate calculus. fractal pairs is another way to turn lagging indicators into leading indicators.

Enter as both show on the 20 (short) or 80 (long) always (last one in)
Exit as both leave the 20 (short) or 80 (long) always (first one out)

Also, are the settings for stoch & macd, simple or exponential?

****Mine look simple; I have never seen any exponential ones unless they are superimposed on an exponential price chart.

Thanks!

kempo

 

[jack hershey]

Hi Jack,

1) You mentioned using MACD (5,13,6) histogram to stay out when there are small ranges. I understand it would be used in conjunction with the stochastics indicator. I'm wondering if I have the correct understanding of how you would use the MACD histogram.

You mentioned the histogram value, ".4". By this do you mean only take long trades when the histogram is > .4 and only take short trades when the value is < -.4?

******Short answer: Yes, exactly.

***Long answer. I am using the histogram here in a simple way that represents divergence of MACD. The "C" pard of MACD. The divergence of the 5, 13, 6 is corrordinated with the Stoc 14, 1, 3. The difference of 13 and 14 is small and either could be shifted a little so let's not bother with this dissenence.

Since the Stoc is relativistic we can't use values like we can with MACD which is an absolute indicator. The point four buiness is a dividing line between flat and trending for beginners.

As we know, you can only make money when there is a trend (up or down). Lateral trends do not make money. The things that happen in lateral trends are opposite pairs: whipsaw and wash trades. A person can get on the wrong side of the micro cycles in a flat low volatility horizontal channel; this is called being whipsawed. On the other hand it is a good exercise to do wash trades for mental proactice in such channels to prove to yourself that you can usually exit any entry with no loss of capital. By seeing that the MACD hisogram is small and smaller you get used to pegging non trends right off and then stay on the sidelines. By staying out of nontrends you achieve a major thing: you are never in the market when it BO's (breaks out) against you.

If a person where always debriefing himself on his efforts, he would find, often, that the market "goes against him". youcan now translate that statement to: "Beginners sit in non trending markets for some reasons, then when the market trend starts up there is a 50% chance that they are on the wrong side of it and they do not have proper protection either". The best protection is to be sidelined in non trends. the second best is to have sequences that tell you trends are coming to ends and to be able to exit far away from your protection and not use protection for exiting. Third best is to be in a place where you an be entering (sidelines are nice place to enter from.

So I am making an effort to get to anticipation. you will see this showing up as your consciousness is raised notch by notch.

Look at mktman a very low consciousness showing there and he has a loud mouth too.

2) When using 5 min bars, do you only take trades upon close of the current bar? Or is it ok to enter/exit trades during an open bar if the stochastics give an entry/exit signal?

For example, say the last 5 min. bar did not give us a long signal, but the current bar is, although the 5 minutes have not completed. Is it ok to take the trade now, or do we wait until the bar closes to see if the entry is still there?

*** this is kewl place to be thinking. We will give it a shot when it occurs and not wait until the end of the bar. Check the 1 min to confirm your entry as you make it. If the trend goes flat just tough it out for a few mins and watch the next bar. If you read bubba7's description of the table and lines you can see the prossessor is just getting a new ink pad.

 

[jack hershey]

This shows a lot of consideration and effort.


How would you suggest to fit the parameters to other markets such as the Euro Stoxx 50 Future? I have done some testing but maybe you have a more clear solution for that matter?
Regarding Stoch, Histogram(Volatility mesurement), Macd


***I am a strong believer in transference of ideas to other market and to other people using the same stuff in the same market successfully.

So go for it. what is extremely important is to keep track of the nuances as you tune it up iteatively. Must markets, today, have a strong PC nd electronic trading influence. This stability of setting is very helpful.

You also can work to find,intially, the best operating fractal. Blueberry cake has a backtester. By using it and not changing the trading algorithm, it has the singular value of telling you where you can make the most money; that is on what fractal.

The beginner stuff is just that. It is a truncated algorithm designed to make money using the fewest parts in an algorithm.

ike we saw money went down by exiting on the first line that left. But look at the number of transactions that occurred: they went from 5,878 to 12,735. The evidence says the thing worked less well because it was a case of jumping in and out too much. It says who cares if the fast line leaves; it is going to pop back much more than half the time. So what do you do to tune it up. LOL you only leave on even fast line exits (the second time it goes out). well you won't make it back to only 5,878 cycle perhaps but you will be in longer and the profits to loss ratio will pick up.

Now if the test were run on an intermediate level where we complicate the beginner algorithm ( Lets say the beginner has confidence and is not escaping trades due of lack of something or other). The first hing we do is not let him into so many trades and we focus more and more on stronger trends by bringing in an absolute indicator along side our relativistic Stoc indicator.

We use the .4 thing with MACD (5, 13, 6). Note that cycles go down and the win ratio will pick up. The profit and loss ratio changes too. Thus as you go into checking out another market you do the same thing. You start simply and add iteratively.

 

[agin1415]

Has anybody tried the Esig Backtesting Function for Stoctastic?

 

[ges]

Quote from jack hershey:


I am going to stay behind the beat for a couple of weeks to find out who can learn what first,

Then I will post 5 to 15 minutes ahead of the market for a while so you can get used to preparing properly for a good shot at some trends.

Thanks for putting it all out there for folks to see and judge. I look forward to following along and esp to your intraday postings.

ges

 

[tampa]

Quote from agin1415:

Has anybody tried the Esig Backtesting Function for stochastic?

You could do that - but you'll do a lot better simply sitting there keeping track of the trades on paper for a few days.

1) You'll reinforce the rules and method.
2) You'll begin to understand what you don't understand
3) Most importantly, you'll begin to believe.

When I first started reading Jack's work, I was convinced that he was a befuddled old man (no slight intended, that's what I am). The next step was to assume that he had something to say, but wasn't very good at saying it. Now I am at a level that truly has me alarmed - I am beginning to understand what he is communicating. Jack sees things a little differently than you or I do. That's what makes it worth the effort to figure out what he has to say. backtesting won't really do that for you.