Current breadth divergence identified using FIRE

16 July

So much for historical examples, what about a current breadth divergence identified using FIRE? Here it is.

The SPDR Consumer Staples (XLP) has 41 components. The chart shows the major breadth divergences between price (top window) and the %age of components above their 200 day moving average (bottom window).

SPDR Consumer Staples (XLP) and 200 day MA breadth

• In March 2003, a new price low forms with a bullish divergence
• In June 2004, a bearish divergence forms. The setback is quite minor and lasts until October. As with XLY at this time, the first divergence in April does not mark the high. Not every divergence is a top
• In December 2007, a bearish divergence  forms. This is a bigger divergence than in 2004 and marks the top
• In April 2010 a divergence forms and a 10 week pullback follows but new highs are formed
• In July 2012, a bearish divergence forms. There is quite a significant drop off for the breadth indicator: from 95% in early May to 73% now. Here is a closer look

• Breadth fell to 61% at the early June low but in the recovery for XLP, the breadth rally has been minimal
• This can occur if highly weighted components are advancing whilst smaller components are falling
• In this case, Walmart (WMT), the 4th largest component, has rallied strongly and Proctor & Gamble (PG), the largest component has rallied since late June

Which stocks are failing?

The following stocks have fallen below their 200 day moving average since early May: ADM, AVP, CAG, EL, HRL, KR, SWY, TAP, WAG.

These are mostly lower weighted components of XLP. You can review those yourself if you like. The exercise of trying to keep track of the market by taking in the action of 100s of stocks in your head is futile. Stick to a top down approach to make sense of things.

Summary

XLP is being held up by a few high weighted components. This is unsustainable and will be followed by at least a pullback, maybe a top. The breadth data does not tell us exactly when that will occur, just that conditions are ripe for a correction. 

Other breadth series that can be calculated using FIRE will be covered in future weeks.