WEEKLY ENGULFING . help needed

sven

Hello

I am looking to create a scan to analyse weekly data and scan Engulfing patterns showing on weekly charts.
It's predefined as daily , but I cannot find the commands to do to it on weekly.. Thanks for the help

markd

It looks like you can just do this:

On the scan workbench, select the the pattern you want from the Candlestick Pattern drop down and add the modifier "weekly" in front of it:

// begin scan

[group is sp500]

and [weekly Bullish Engulfing is true]

rank by PctRelative(189,$SPX)

// end scan

The last line is optional. It puts the results in order by their relative strength vs. the sp500 index over the last nine months (189 days if you assume 21 days in a month; you can change that number to your preference; however, PctRelative doesn't seem to work on a weekly time frame, so you have to use a daily parameter, e.g. 30 for six weeks (5 days times 6 weeks).

For more on modifying expressions, see this page, about half way down (the rest of the page is interesting, too):

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

sven

Thanks a lot Mate !

sven

Mark thanks a lot , it works very nicely . I have another question for you. I am looking for a scan to analyse on daily wolfe wave or the formation 3 indians (Up trends of Down Trends.) Did you see something like that from the community, it's quite hard to modelize .. Cheers

markd

I don't think I can help you there. The scan engine is really designed to find state changes, or combinations of state changes, like crossovers or breakouts - e,g, MACD Line x Signal, Fast Stoch x 20, etc., and filtering for states in effect, e.g. close > 200 MA. It's not so good for patterns wider than a few bars. If you are very patient and like to experiment, you can sometimes find simple patterns, but not without missing some instances or catching things that are not what you intended. Market action is highly variable, but scans have to be specific to some degree, and there will always be patterns that you would choose by eye, but fall just inside or outside the limits you write into your scan. And that doesn't take into account how subjective the process of recognizing a pattern can be in the first place.

But maybe some one else here knows where you can find something useful.

sven

thx a lot Mark and happy new year