Adx Scans

Sofearnot

Trying to figure-out how to Search for Minus DI Within +/- of the Adx Sig. Is this possible? Not Sure.

Also, what exactly is the Difference between MACD Line and MACD Signal?

Sofearnot

Within 5% or 10%

lmkwin

DI- < ADX * 1.05 // DI minus is less than 5% above the ADX
and DI- > ADX * 0.95 // DI minus is greater than 5% below the ADX

in other words, DI- is within +/- 5% of ADX

Link below discusses scan clauses for proximity to and for near crosses. The examples and sample clauses are using price, you would just change the price to be the terms you want.

https://support.stockcharts.com/doku.php?id=scans:advanced_scan_syntax:near_crosses


macd 12,26,9. The 12,26 are the moving averages used to form the LINE, 12ema minus 26ema = line value.
The 9 is the moving average of the Line and is known as the SIGNAL. So the Line will lead the signal in both directions. The Histogram is the 12-26 value. If the 12-26 > 0 histogram is positive. If 12-26 is < 0 histogram is negative. macd deals with points or dollar value. So a high priced stock will have a different scale on the macd compared to a low priced stock. So if you scan for macd line > 10 you will get a certain price segment of stocks that can have $10 size of macd. Using PPO instead and PPO line > 10 wil show you securities that are 10% above the zero line so it's more universally able to be applied.

Sofearnot

I knew my Stockcharts.com bretheren would find a way! ha I had actually been Reading, quite a bit, and couldn't figure it out. Knew it had to be Something simple, I was missing. you guys pulled through, Again.

On the stocks page, it only has Hist (Which i know), and MACD (generically). i am assuming that the BLACK line must then be the Line and the RED must be the SIGNAL, based on what you said. It combines the two together to just "MACD."

Have always used Macd, even tho I know the PPO is a bit of an improvement- only because I am looking for specific Patterns that I know happen on Macd, it now appears that the patterns may be Identical, or very similar between the two (I think I had different numbers I was using for each, before) So, maybe I'll start using the PPO, instead, now that I am using matching numbers

Sofearnot

question, is there a such thing, as a "between" parameter? .."Is between..." ?

i.e. macd or ppo "is between 5 and 10" ... I think you just have to Write it out, with two steps, as in your Above example. Is that true?

Is there a 'Suggestion box" anywhere

Sofearnot

...Not that using Two steps is all that complicated...but, if you do that for Too Many parameters it Could potentially introduce some either human or computer Error.

Sofearnot

If not, No worries.

Sofearnot

not needing anyone to work out the PPO example above, just Wondering if it Can be Accomplished in 1 step versus two.

lmkwin

There is no between. > < = would be math.

X is a cross above. There is no cross below. For cross below you reverse the statement.

close x sma20 for close crossing above sma20.

sma20 x close for sma20 crossing above close

You could also use and close > sma20 and 1 day ago close < 1 day ago sma20 for a cross above. Reverse the > and < for the cross below

Is and Is Not for text type of filter. Group is Industrials. Group is not industrials

Sofearnot

Okay, what would be Looking for a PPO crossing below 0, Within the last 10 days?

lmkwin

min(10,ppo line(?,?,?)) < 0 // the min is the lowest the ppo has been in the last 10 days is less than 0
max(10,ppo line(?,?,?)) > 0 // the max is the highest the ppo has been in the last 10 days is greater than 0

So the line has been above 0 and below 0 in the last 10 days.

If you want a line that was above before and has moved down in the last day to two or three or.... move the max back a few days and shorten the min period as well.

min(3,ppo line(?,?,?)) < 0 // min is the lowest the ppo has been in the last 3 days is less than 0
2 days ago max(10,ppo line(?,?,?)) > 0 // the max is the highest the ppo has been starting 2 days ago in the last 10 days is greater than 0

The above will allow the PPO to have been above 0 in 10 days starting two days ago. Today is the 25th so it would be looking at the 10 day period ending on the 23rd. And the ppo has been below 0 some point in the last 3 days.

if you need today's ppo to be less than 0 add

and ppo < 0


To answer a prior comment on this thread

"Is there a 'Suggestion box" anywhere"

Yes, you can submit a request to StockCharts.com Support. If they can do it they often will. If they can't they won't. There is a thread on this forum about making requests to Support. I'd have to find it to see if "we" got anything out of those.

https://scan.stockcharts.com/discussion/1721/suggestions-to-stockcharts/p1

Yes, it does look like we got some suggestions implemented. So make your suggestions to StockCharts support and copy the request before you submit it, and paste it to the forum thread for others to see and possibly also suggest to StockCharts.com support. The more making similar suggestions, the better the opportunity to see it happen.

Sofearnot

Okay, thanks buddie Good stuff.

Sofearnot

Something's not working, i think. I try and write a scan :
[type is stock] and [country is US] and [max(5,CCI(20)) < 0] AND [CCI(20) > 0]

I find it very hard to believe that there are *NO* stocks Where cci has gone positive For the First time in 5 days?? Out of the Whole universe of 5,000+ stocks... i am thinking this Can't possibly be right. What am i doing Wrong here??

Sofearnot

Even went back each of several days, Absolutely no Hits.

markd

You could try just

and [0 x CCI(20)]

or, instead of looking for the specific day of a crossover, you could specify a range

and [min(10,CCI(20)) < 0] AND [max(5,CCI(20)) > 0]

or something along those lines.

With either approach, set the date back, say, 60 days and come forward, either a day at a time, or five days at time, etc.

But, if you look at a chart of say the SPY, crosses above zero occur only once or twice a month at most. Since most stocks move more or less in tandem with their index, you might cluster your searches around a date when a relevant index or ETF based on an index (XLE, XLB, etc.) crosses over.

lmkwin

[type is stock] and [country is US] and [max(5,CCI(20)) < 0] AND [CCI(20) > 0]


and [max(5,CCI(20)) < 0]
You are asking for the Maximum or Highest value of the CCI20, in the last 5 days to be LESS than 0

AND [CCI(20) > 0]

This is the current CCI20, which is also included in the Max5 period to be GREATER than 0.

Asking for CCI20 in the last 5 periods to be less than 0 and the most recent value of the last 5 days to be greater than 0. You will get 0 results.

You'd have to push the Max period back a day or so to get the current period out of that max period.

1 days ago max < 0 and current period >0

markd

Wow, I'm slipping. I should have caught that.

Sofearnot

A little simplified version, of what I'm trying to do, (I didn't want to lose anybody.) Wow, that works! I added the "one day ago' and got 40 hits!

Sofearnot

@markd No, you are not slipping at all. I would put your accuracy, at better than 90%

markd

Right, that's slipping.

Sofearnot

Haha Lol.

Sofearnot

Haha, We all must be slipping then

Sofearnot

i didn't catch it. but, i never professed to be expert at writing scans...

Sofearnot

Thanks @markd and @lmkwin

Okay, so I'm trying to make a Scan that most approximates this pattern: 200 day descending by a certain percent/gradient, macd Was above 0.50 recently, but now is Just under zero within a few days, not necessarily today though. The price And/Or the short term moving average was parallel to the 200 day for a short period of time. Trying whenever possible to not using exact price so as to capture more stock hits. So essentially the price/m.a. pattern and the macd (or ppo). It's only two things, but my scans are not getting any hits; maybe i'm using too many variables? Is this do-able?

lmkwin

It's much easier to assist if you post your scan code in question for assistance.
What chart is that one? My BWEN chart looks nothing like that.




Parrallel is going to be an issue, if you consider that the different length MA's, or Price vs MA move at different pace or stride. You could look at the PctDiff between the two at different points along your "short period of time" interval. The X days ago PctDiff is +/- Y% at the beginning of the short period of time X days ago. It is the same +/- at the middle of the short period of time Z days ago. Z is less than X. It is also +/- the same at the end of the short period of time.

The MACD should be 'simple' but you would need to pick your levels

//MACD line was above 0.5 recently and just under 0 within a few days.
and [max(10,MACD Line(14,17,9)) > 0.50]
and [min(3,MACD Line(14,17,9)) < 0]
// MACD Line is greater or equal to the below zero value.
//If it needs to be lower than the min then reverse > to <
and [MACD Line(14,19,9) >= min(3,MACD Line(14,17,9))]

Sofearnot

yeah, ur right must have got things mixed-up when I condensed everything onto Word document. Seems like it could work. Going to focus on Longs since i do Believe the manufactured Crisis with the $40 trillion Debt Ceiling is going to be "resolved" and wait till the next 'Crisis'

Sofearnot

the Next generation gets to deal with our own Manufactured problems : )

Keep the Party Going...!

Sofearnot

They're Going to keep doing this 'Debt Ceiling" Shenanigans - Regardless of political party - Every Year or Every other year- Until it comes to Pass that social security recipients don't get anything...then, there will be the desired outcry if even too late. Well, That's what happens when you have a congress made up of Lawyers and 0 economists i guess. Whoevers reading this Post in five or 10 years is going to Think i'm Absolutely Psychic & have magical powers.

Sofearnot

Another 1.4 Trillion of Inflation politicians all Patting themselves on the Back now at this success

Sofearnot

If we have a government that has zero dollars that Overspends 10 trillion dollars every four years somebody explain this to me what is the utility of this

Sofearnot

Lol i just don't think it's Appropriate for our Elected leaders to be Spending all our money Decades b4 we Even Earned it