[grbaNT]

Yep, had the exact same feeling. Was expecting up week in the Survey. That is why I like the market breadth. Waiting for 3% and bigger weekly moves in the averages and the survey out of the range is the real signal to put much more cash at work. Best

I know you don't use the NYSE Percentage of Stocks above their 150 day MA chart, and the other time frame MA breadth averages due to your survey, but the daily version of the 150 day MA does give some extra context to the week, as it was almost 7% lower than last week at it's low, but although it finished the week lower than the previous by 2.53%, it did have a recovery near the end of the week to finish higher than it opened the week, and so recovered over 4%, and is also around 10% higher than it was at the April lows a month ago, so there is a fairly strong divergence showing in the NYSE Percentage of Stocks above their 150 day MA chart. And also the other short and longer term timeframes too. The short term especially, which gave a new P&F Bull Confirmed signal on April 10th, which was then followed by the 150 day chart on April 16th. (See http://www.investorsintelligence.com/x/b...ators.html for descriptions of the P&F market breadth signals for anyone reading this that doesn't understand what they are).

So I find the usefulness of the multi time frame approach from these missing from Weinstein's survey. Whereas the moving average breadth charts can give short, medium and long term information, as well as being used together as one for the weight of evidence approach. Hence I think there's benefits to both the stage survey and MA breadth if used in a multi time frame approach for the weight of evidence. As have been clear leading indicators over the years.

[quote pid='12107' dateline='1525759278']
There is one other market breadth indicator I use that I would like to share here if you agree. I use the same sample as for the Survey but the data set is smaller as I started putting it together almost two years ago. Stock classification is done as per Chuck Dukas's TrendAdvisor Diamond described in his book. ...

Interesting, I added TrendAdvisor screening to my database here.    It seemed fairly easy to classify stock programmatically. However, I did not have much luck the time I tried it ('paper' trading I think).  So I did not take it further.  Perhaps I should have stuck at it longer?
[/quote]

It is easy, it is just the question of what platforms you use. I use NT's Market Analyzer to extract data and just paste them to excel sheets that do the rest of the work. Yep, fairly simple! You can do this every day, just requires some extra work after hours.