Due to some interesting discussions on TW concerning the current correction's tendency of price leading internals to the down side thus far, decided to take a historical look at the movements of internals versus price during market corrective phases. This study uses the NYSE breadth and volume cumulative lines as the internals' proxy, and the SPX as the price proxy from 1940 through the current market climate.
When considering approaches to such an exercise, I quickly realized filters would need to be applied for comparing historical events similar to the recent market conditions:
1) The SPX must have made at least a new six month price high within the prior two months.
2) The SPX cannot have endured a correction of over 5% within the prior
3) The data points under examination are the number of RAAD line points, number of RAUD line points, level of the NYSE AD McSum, number of days since the six month high was posted, and the SPX percentage loss or gains at arbitrary time frames going forward.
4) All of the data points described in criterion (3) are captured on the day the SPX has declined 5% from its six month high.
A goal of the exercise is to compare the number of points the ratio-adjusted AD (RAAD) and UD (RAUD) lines declined from their respective six month highs when the SPX declined 5%. On a closing basis, the SPX has not yet declined a full 5% from its late February 2007 highs, but has given up a little over 5% in that time frame when considering the intra-day highs and lows.
Since 1940, there were 28 events fitting the criteria, with 19 of those events occurring since 1960, which is the basis of this study.
The following table provides the various statistics on an event basis since 1960, but includes the average statistics of the qualified events from 1940, 1950, and 1960. The qualified events since 1960 have been sorted, in ascending order, pivoting upon the number of NYSE cumulative RAAD line points lost from the SPX six month price high until the day the SPX took at least a 5% haircut.
I was surprised to see the most recent decline required the fewest number of RAAD line points, of all the events under consideration, to accommodate a 5% price decline for the SPX. Five of the events required fewer RAUD line points to be sacrificed in achieving a 5% SPX price decline than the current correction.
Therefore, on a historical basis, there is not much question price is leading the internals in this most recent price correction. The qualifying event on March 27, 1997 did not require many RAAD or RAUD line point sacrifices to support a 5% SPX decline, so the bulls need to keep that in mind after examining the SPX price returns over the next ten days following the initial price decline in 1997. Also note the 5% decline from the six month price high in 1998 took 8 trading days... the same number as the current event, although the 1998 decline required far more internals line points to get there.
There is noticeable variance in the price action going forward after a qualifying event, so as usual, there is still a non-zero probability of further price deterioration over the coming days/weeks.
The bottom four rows of the above table are derived from adding yet another filter for a qualifying event: the AD McSum must be greater than +500 on the day the SPX achieved its 5% correction. Including the 2007 qualifying event, there have only been four events meeting the required criteria, and the prior three were all posted before 1960. The 1940s in particular were a very volatile period with respect to price and internals with many extremes surfacing in the McSum levels in those days.
Therefore, along with the very small sample space, I am not giving the results of those three events that occurred with an AD McSum over +500 a lot of weight, but the performance of the SPX going forward for each time frame was impressive, giving credence to the notion an AD McSum over +500 is a formidable hurdle for the bears to drive prices down significantly under such conditions.
Food for thought and FWIW,