As you know, TC2000 offers Heiken-Ashi charts.  These are very useful for showing persisting trends but use calculated rather than actual prices.

I would like to create an EasyScan that would alert me to incidences in a group of securities where there have been multiple periods where a security has closed the bars of a Heiken-Ashi chart at a lower price than it opened.  In other words, base the EasyScan on the derived Heiken-Ashi values rather than the actual values of the period.

Can you help me with this?

Dennis

Hi Bruce,

The charts that I use for trading are quite short term, three minute bars.  The scans that I use work very well, no complaints whether they are created from the charts or from PCFs. 

But ...  The maximum time the alert will remain in view is three minutes and the minimum time can be as little as a few seconds if the alert pops up just before the bar expires.  And sometimes, with buying and selling, etc, the alerts disappear without having been seen or acted on.

Some questions:

1. How can I create a log of scan alerts in chronological order, either in a flat file or a watchlist?

2. How can I use a scan to detect, as a simple example, a macd crossing an ema signal line in the prior three minute bar (my main scan alert) and still remaining above it in this three minute bar?

I think I can do item 2. above in a PCF if all else fails:

XAVGC18 - XAVGC40 > XAVG(XAVGC18,4) - XAVG(XAVGC40,4)
AND
XAVGC18.1 - XAVGC40.1 > XAVG(XAVGC18.1,4) - XAVG(XAVGC40.1,4)
AND
XAVGC18.2 - XAVGC40.2 < XAVG(XAVGC18.2,4) - XAVG(XAVGC40.2,4)

This seems to give me what I want in the sense that it will keep the alert on the screen a little longer when I use this script together with my chart scan which looks at the current period only.  I know just a scan is preferable, but I don&#39;t know how to reference the current period AND the prior period in a scan where the trigger is the crossing of an ema.

My PCF above seems to be doing what I want, but I would appreciate your comment on whether it is the best way to hold the alert.

Thanks in advance for any help you can offer with these questions and in prolonging scan alert visibility in general.

Hi Bruce,

You said in this response on Apr01/14:

By extension, suppose I had the Pre/Post Market Data set on my 5 min chart at &#39;Always Show&#39;, and I had a streaming scan to alert me to a macd bottom.  And suppose there was a major gap opening in, say, GLD, or one of the active ETFs, there could be an apparent contradiction between the macd bottom of my chart and the alerted macd bottom of the scan. 

Both the chart and the alert are &#39;correct&#39;, but they don&#39;t agree as the base data is different.

To me, the apparent gapping is best understood by including the pre/post data but this understanding is blunted by the timing change on the alert.  And in days following, with the pre/post no longer included in the base data, only the &#39;alert&#39; macd bottom is reflected for analysis without the &#39;smoothing&#39; effect of the pre/post.

Any resolution in the near term, or am I best advised, during the early part of the trading day, to continue to mentally adjust my reaction to the alert based on the extent of a gap open?

Thanks in advance for your answer, and for all of your other answers to date.  All have been interesting and informative, and some are examples of truly innovative solutions.  Very well done.

HI Bruce,

OK, let&#39;s see if I&#39;ve got it right.

I have created an indicator, linked to &#39;Price&#39;, identified on the chart as &#39;FAvg 10&#39;. 

I have created a second indicator, linked to &#39;FAvg 10&#39;, which is a one period EMA of the &#39;FAvg 10&#39; indicator, and has the Period set to &#39;1&#39; and the Offset set to &#39;2&#39;.  This indicator is identified on the chart as &#39;EAvg 1 +2&#39;

Then I created a scan condition for the &#39;FAvg 10&#39; indicator, which has as its Condition &#39;Crossing Up Exp Moving Average 1 +2&#39;.

This scan appears to give me the results I want.  Does the above construct, to generate a bottoming ema, appear correct to you?  You suggested " a 1-period ema with the Offset set to 1 moving down.".  I used an offset of &#39;2&#39; because I wanted to pick up a value two periods ago, but I&#39;m not certain that I&#39;ve got it right.

On a general note, is the purpose of &#39;offset&#39; in the moving average toolkit to give us a chance to do these multiperiod compares?  Where can I read more about the purpose and usage of the offset value, both positive and negative signs&#39;?

Hi Bruce,

I need a bit more clarification here..

Suppose I needed to create a new scan that alerts me to a bottoming 10 period FW ema.  I want the current period ema to be greater than the prior period, and the prior period to be less than the one before.

This is easy (but slower) in a PCF, but how do I do it using Create Scan Condition methods?

I need two conditions in the scan.  The first one uses the condtion  of &#39;Moving Up Net&#39; with the Period equal to &#39;1&#39; and Net Change set to &#39;0.00&#39;.  That gets me watchlist items with the ema moving up this period relative to the prior period.

The second condition, I think, would use the condition of &#39;Moving Down Net&#39; with the Period equal to &#39;2&#39; and Net Change set to &#39;0.00&#39;.  But I suspect that what I am getting is condtions where my current ema value is less than the ema value of 2 periods ago.  That is, 0 > 1 AND 0 < 2.

What I want is 0 > 1 AND 1<= 2, given that &#39;0&#39; is the current period, &#39;1&#39; is the prior period, &#39;2&#39; is the period before that. 

How do I do that without being able to specify the period numbers to compare, using the more efficient  &#39;Create Scan Condition&#39; rather than a PCF?

Hi Bruce,

I enter my long positions in a personal watchlist, then have a scan focused on that watchlist to detect sell conditions so that I don&#39;t have to focus on the positions until necessary.

Works well, but ... the scan does not see the frequent changes in the &#39;Owned&#39; watchlist unless it is refreshed by switching from &#39;Streaming&#39; to &#39;Scan&#39; and then back to &#39;Streaming&#39;.

This is not a big deal, but I thought I would ask  -  Is there any way that I can automate the refresh process?

Thanks

Hi Bruce,

I enter my long positions in a personal watchlist, then have a scan focused on that watchlist to detect sell conditions so that I don&#39;t have to focus on the positions until necessary.

Works well, but ... the scan does not see the frequent changes in the &#39;Owned&#39; watchlist unless it is refreshed by switching from &#39;Streaming&#39; to &#39;Scan&#39; and then back to &#39;Streaming&#39;.

This is not a big deal, but I thought I would ask  -  Is there any way that I can automate the refresh process?

Thanks

Hi Bruce,

Great, that&#39;s clear in my mind now and I will use &#39;Create Scan Conditions&#39; wherever possible.

So I can use multiple &#39;Create Scan Conditions&#39; for the same scan, but they will be cumulative with an expressed or implied &#39;AND&#39; between them - but &#39;OR&#39; cannot be used.  Correct? 

My scans are never complicated, but there is one scan I have never managed to create in &#39;Create Scan Conditions&#39;.  That is to find instances where the 4 period ema of the 12 26 4 histogram is greater than the prior period, ie a rising ema (or macd or whatever) line.  Ideally, a bottoming ema, where this period is rising but the prior period was falling, but that may be asking a lot.

I do this now using a PCF, and it works, but the speed is &#39;Slow&#39;.   Does &#39;Slow&#39; imply that the results of the scan will be delayed on receipt by, what,  - 1, 2, 5, 20 seconds?

With reference to the ema question above, how do I compare one period of an indicator relative to the prior period of the same indicator in a created scan condition?

I see conditions like &#39;down net&#39;, &#39;down %&#39;, &#39;down amount&#39;, &#39;down smart&#39;, but nothing &#39;down relative&#39;.  What am I missing?

Thanks

Hi Bruce,

I have learned to write any PCF that I have need of, and have recently used the &#39;Create Scan Condition&#39; feature on your charts for comparing related indicators - with good results.  Are these two scan methods equal in their results and speed of execution?  Any reason to favor one over the other (ie my PCF vs your internal code) for regular use in streaming scans?

Thanks

Hi Bruce,

I know that I can compare a price (open, hi, lo, last) to another price that is, for example, the highest value of the last 24 values.

Such as   ((MAXH24 - C) * 100) / MAXH24 > 1.0   to find a symbol that has dropped by over 1% from the high of the last 24 periods.

I would like to do the same type of scan for the ema, macd, histogram, but cannot find a way to make the MIN and MAX tools work with the XAVG tools.  For example, a symbol trading with the macd at the lowest value in the last 24 periods.  I&#39;ve tried several ways, logical and illogical, but I only get Formula Errors.

Please help.

Dennis