Introducing breadth data

07 May

This week I will show my findings from some breadth data that I have got Pinnacle Data Corp.

Breadth indicators show the number or percentage of stocks in a group that are going up. The criteria for ‘going up’ is always a calculation, not an opinion.

At stock market peaks, breadth indicators show an index uptrend weakening through lower stock participation. The index makes a new high but the breadth indicator does not: there is a divergence.

Things are different at lows but given the rally since spring 2009, the current focus is on whether there are any breadth divergences that suggest a market top.

You can use whatever you like as your criteria for ‘going up’ but there are some common measures.

• %age of stocks in a Point and Figure uptrend
• %age of stocks above a moving average

These are the main two that can be targeted at any group of stocks you like. You can vary the criteria of each to show long and short term breadth conditions.

The data from Pinnacle is broad market data, it includes all stocks on an exchange, even things that trade as stocks but aren’t really stocks (like bond funds).

For the NYSE, NASDAQ and AMEX there is:

1. The number of advancing issues (stocks) and declining in issues.
2. The number of new 52 week highs and new 52 week lows.
3. The total volume of advancing issues and declining issues.

Data is provided for these each day and for some, on a weekly basis. The weekly basis is not the sum of the days. You could have four days of fractional gains and one big down day, a net of three advancing days but one declining week.

Data is plotted cumulatively: add the next day to the running total. Some data goes back to the 1960s, much to the early 1980s.

What is the predictive track record of this data in predicting tops?

• The new highs/ new lows data has the best predictive record
• The cumulative new highs/ new lows lines are smooth, they have less noise, so are easier to interpret
• This data works well on all of the NYSE, NASDAQ and AMEX but best on the NASDAQ stocks
• The other two measures on the NASDAQ have poor track records
• The advance decline data has some use on the NYSE but doesn’t help at all on the NASDAQ and AMEX
• The advancing and declining volume data has a patchy record

These observations apply across the data and weekly data. I will show the current observations in a separate posting.