A rising market (i.e. stock index) is usually characterized by rising price and wide breadth i.e. the vast majority of stocks in the index are increasing in price. When price and breadth of the index diverge, it indicates the foundation of the index increase in price is starting to crumble. Price weakness is expected to follow, so that breadth can be viewed as a leading indicator for market weakness. This proved to be the case earlier in 2015 where breadth began to diverge from price in ~May until we ended up with the Flash Crash in August. Below are 2 examples using the NYSE where indicators of breadth and price diverged between mid-April and mid-May. Breadth decreased and price followed in June and on into the summer where the divergence disappeared.
In the past few weeks, focus has returned to breadth and the perception has been that we once again have a divergence, which may portend another price correction. Generally speaking I believe the analysis of breadth is more nuanced than is generally appreciated, and so I thought it might be a good idea to have a look at recent market breadth to see if there really is a divergence. I have chosen to examine range of breadth indicators to provide as clear a picture as possible. As you will see below, there is very little evidence to support the contention that we currently have significant divergence between price and breadth of the major stock market indices.
The indicators I chose to represent market breadth are:
1. Number of stocks above the 200 day moving average - this is a long term view of market breadth, but similar conclusions can be drawn using the short-term 50 day moving average.
2. Number of new highs - new lows. The theory is that a rising market will be characterized by a (relatively) high number of news highs, and a descending market a high number of new lows. In reality during an upmarket the number of new highs tends to saturate, but in a down market the new lows rise quickly giving rise to spikes in the indicators. This indicator can also be cumulated over time to track an index.
3. The advance/decline line of the index. This is classic breadth indicator, and is calculated by cumulating the daily number of advancing minus declining stocks for an index.
4. The Hi-LO indicator.
5. Equal weight index calculation vs the standard market capitalization weighting. By equal weighting the index, the impact of smaller stocks is increased since the index price is no longer dominated by a relatively small number of the largest capitalization stocks. This reveals the influence of the large cap stocks i.e. it indicates index breadth.
6. The summation index. This is an oscillator that is calculated from moving averages of the A/D line.
All the indicators are readily available on common chart plotting packages. The examples I provide here are from Stockcharts.com
Number of Stocks Above the 200 Day Moving Average
The 200 MA gives a long-term view of the indicator , and in this case we use the SPXA200R indicator. As shown below, this indicator showed a divergence with price starting in May until mid-July, after which the divergence disappeared. At that point, however, the indicator was at a low value of 55% while the index had remained approximately stable. The issue, then, was not the divergence of the indicator and price, since this divergence had disappeared by mid-July, but the low value of the indicator. The drop of SPXA200R relative to the index was a warning sign, and the follow through occurred in August with the flash crash.
Once the crash had occurred, and since that time to the present, the SPXA200R indicator has tracked the index price closely - see below. In other words, there is now no divergence between indicator and price. The indicator is at the relatively low value of 48%, but this is not inconsistent with the price of the index, which is now well below it's all time high.
Relatively low values of the SPXA200R breadth indicator are in fact consistent with values that were observed during other recent market corrections in 2010 and 2011, and in both of these situations, the market went onto subsequently rally. These observations have been discussed in detail in a recent article. One cannot conclude from these weak indicator values that they are pointing to a major market correction at the present time.
Number of New Highs-New Lows
The 2 charts above show the New Highs - New Lows percentages for the SPX and NDX. There is clearly no divergence in direction for both indices: price and breadth and moving down. Below is the indicator for the NYSE shown as a cumulative, which displays no hint of divergence with price.
And just to point out the nuance of breadth analysis, it is important not to mix breadth indicators between indices. I've seen this done within the last week in several relatively high profile articles. The example below shows the SPX compared to the NYSE cumulative new highs - new lows; it would be straightforward to conclude there is a divergence that is simply not there if the analysis is done properly.
We can conclude that New Highs - New Lows do not indicate a divergence for SPX and NDX.
Advance / Decline Line
The A/D lines for both SPX and NDX have not been able to successfully identify divergences all year long for SPX and NDX. A divergence did appear for both indices between mid-May and mid-July and then the divergence disappeared. Not much good in terms of signaling a significant market drop in August.
This indicator tends to be pinned at 100 during rising markets, and shows dips as price drops. For SPX it could be argued there was a divergence between mid-May and Mid-July, but since it then disappeared it did not portend the August drop. Not much use. Both SPX and NDX show no current divergence.
Equal Weight Index.
The same story here at above: no support for a divergence in the most recent data. Some indication mid-May to Mid-July, but then the signal disappeared.
This indicator appears to be very sensitive to breadth, but even so the most recent data for NYSE shows no divergence. It's interesting to see that there was an obvious divergence in the May to June timeframe with negative breadth and approx. stable price. Then price decreased from mid-July to just before the Flash crash while breadth improved. So actually 2 divergence signals here. A similar obvious divergence is evident for the NASDAQ (COMPQ) and this signal, if viewed across the May- August timeframe, did not disappear. (I wish I had done this analysis in July).
So, what do we extract of value? Well, it has to be said that it's difficult to have solid conclusions with such a small sample size of observations. Nevertheless, some useful information can be gleaned. Most measures of breadth don't appear to consistently signal imminent market declines. Of course, declining breadth usually accompanies declining price, but what I was trying to do here is determine if breath is a reliable leading indictor. It does, however, seem that the summation index is particularly sensitive, making divergence more obvious. It's perhaps worthwhile to pay attention to this indicator as it moves below zero.
I am unconvinced that the most recent market data indicates that breadth is signaling a major market decline via divergence. Decreasing breath is not a welcome sign, but so far there does not appear to be an indication that is anything but the run of the mill decline that accompanies movement of the market indices. Indicator levels at the end of last week are getting close to, if not already consistent with, previous levels where short - term market bottoms formed. This suggests we are probably going to see the market pick up in the next few days, unless the oil price situation deteriorates further, or the high yield market suffers and even more serious blow-up.